Evaluation Model of Football Players’ Training and Teaching Actions Based on Artificial Intelligence

Yun Feng; Yongan Wang

doi:10.1155/2022/7427967

Evaluation Model of Football Players’ Training and Teaching Actions Based on Artificial Intelligence

Feng, Yun;Wang, Yongan 2022-08-30 00:00:00 Hindawi International Transactions on Electrical Energy Systems Volume 2022, Article ID 7427967, 11 pages https://doi.org/10.1155/2022/7427967 Research Article Evaluation Model of Football Players’ Training and Teaching Actions Based on Artificial Intelligence Yun Feng and Yongan Wang College of Physical Education, China West Normal University, Nanchong 637002, Sichuan, China Correspondence should be addressed to Yun Feng; fengyun1991@cwnu.edu.cn Received 10 June 2022; Revised 9 July 2022; Accepted 18 July 2022; Published 30 August 2022 Academic Editor: Raghavan Dhanasekaran Copyright © 2022 Yun Feng and Yongan Wang. *is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Football is a sport that needs to combine the physical stamina and physical characteristics of athletes. It needs to pay attention to the diﬀerences in diﬀerent individuals and then conduct targeted training. In this regard, this article introduces artiﬁcial in- telligence technology into the teaching actions of football player training and analyzes the teaching elements according to the characteristics of the football player’s body. *rough the analysis of fog computing under artiﬁcial intelligence, this article aimed to study the related eﬀects of combining intelligent technology on the basis of athletes’ original training. *is article proposes the establishment of a system model and quantitative analysis of diﬀerent frames of football players’ movements. According to the combination of two-phase analysis, it can be concluded that after the introduction of artiﬁcial intelligence technology, the ability of football players in various indexes has increased by 20%. more eﬀective contact methods during training; on the other 1. Introduction hand, it provides more diversiﬁed training methods for With the development and application of computer tech- athletes in future training, so that players have higher in- nology, artiﬁcial intelligence technology has been widely terest and enthusiasm in training, so that athletes can be used in sports training. Under the premise of improving the more professional and precise in football performance. At competitiveness of football, it is very urgent to seek to the same time, it provides a certain theoretical basis for the improve the level of training. Observing a lot of training data sports group and opens up new research points for the study can show that the commonality of basic training needs to be of football player training and teaching. Based on the above strengthened. *erefore, in the context of the improvement related content, Hassabis et al. found that the ﬁelds of of the global football level and the diversiﬁcation of the neuroscience and AI have a long history [1]. In addition, training process, the lack of personalized training methods Raedt et al. believed that intelligent agents interacting with and content has been unable to adapt to the trend of football the real world will encounter individuals, courses, test re- sults, etc., need to reason about the attributes of these in- development. *is method will aﬀect the improvement of athletes’ abilities to a certain extent, and it is very necessary dividuals and the relationship between them, and deal with to introduce artiﬁcial intelligence technology into football uncertainty [2]. At the same time, Rongpeng et al. believed training. that 5G is a key enabler and infrastructure provider in the On the one hand, by visually expressing the training ICT industry by providing various services with diﬀerent content of the vertical and horizontal movements of football needs [3]. In addition to the abovementioned views on AI, in the form of data, the research is conducive to the im- *rall et al. believed that it is driven by the availability of provement of the training ability of football telemobilization large data sets (“big data”), signiﬁcant advances in com- under the condition of artiﬁcial intelligence. In turn, it puting power, and new deep learning algorithms. Global provides a theoretical basis for football coaches to provide interest in artiﬁcial intelligence (AI) applications, including 2 International Transactions on Electrical Energy Systems imaging, is high and growing rapidly. In addition to de- latency, and fog computing has a vast geographic distri- veloping new AI methods themselves, they are also faced bution and a large-scale sensor network with a large number with better ways to share image data and standards for of network nodes. At the same time, through the connection validating the use of AI programs across diﬀerent imaging between the fog node and the device, fog computing can platforms [4]. Since this article mainly introduces artiﬁcial reduce the processing burden of resource-constrained de- intelligence research, Glauner et al.’s main research direction vices, meet the requirements of delay-sensitive applications, is to use artiﬁcial intelligence (AI) to solve diﬀerent prob- and overcome the bandwidth limitation of centralized lems. Finally, it investigated these research works in a services. *e basic framework of fog computing is similar to comprehensive review of the algorithms, features, and data cloud computing, but its lower-level architecture has special sets used [5]. In view of the fragility of the above content, and components that are sensitive to time response. With this in view of the great interest in artiﬁcial intelligence by feature, you can further control and enhance services such as Seyedmahmoudian et al., it links football and artiﬁcial in- sports. Figure 1 shows a mode of fog calculation. telligence technology together. Extensive research has been In Figure 1, there are three layers as a whole: cloud layer, conducted on various ways of football. According to the fog layer, and user layer. Combining the fog calculation with robust, reliable, and fast performance of the artiﬁcial in- football sports can provide timely feedback of sports-related telligence-based MPPT method, the motion posture of the data, thereby forming a benign closed loop. *e cloud layer human body is discussed under various conditions [6]. is a computing cluster composed of large servers, located at Wang and Liang obtained the upper limb trajectory in 3D the top of the entire architecture. *e fog layer is mainly space and realized the upper limb trajectory extraction [7]. composed of various embedded computing devices, located Because artiﬁcial intelligence and algorithms are inevitably between the user and the cloud, mainly provides a de- inseparable, Patel and Savsani proposed a Stirling multi- ployment environment for delay-sensitive applications, and objective optimization strategy and introduced the multi- responds quickly to user requests. According to the physical objective TS-TLBO algorithm to get the real Pareto frontier. distance between the fog layer device and the user, the fog By comparing the optimization strategies between diﬀerent layer device is subdivided into multiple sublayers, and the closer to the user side, the lower the network delay of the algorithms, the relevant research methods of artiﬁcial in- telligence are proposed, and the best combination with the sublayer. *e network communication overhead between teaching design of football players is obtained and then put devices in the same sublayer is signiﬁcantly lower than the forward the relevant model establishment formula to cross-layer network communication overhead [1, 9]. *e quantify the athlete’s movement posture [8]. user layer directly interacts with the user, sends the collected *e innovative points of this article are as follows: (1) data to the upper layer, and receives the returned result after computer technology is used to establish a systematic study the upper layer processing, as shown in Figure 2. of the actions of football players. Personalized research on *e fog is located between the terminal device and the athletes with diﬀerent personal information is conducted, cloud and builds a bridge between the terminal device and collected system management information is managed, the cloud. Fog computing, like cloud computing, can player training content is recorded, and model analysis on provide computing, storage, and network services for the collected information is performed. (2) *e constructed terminal devices. Unlike cloud computing, fog computing network model is used to design and develop diﬀerent is closer to terminal equipment. *e main characteristics of teaching systems. *rough this system, it can help the fog computing are shown in the ﬁgure: short-distance, analysis of football players to quickly understand their own distributed deployment makes fog computing have many advantages and problems in the process of sports, so as to advantages, including low latency, location awareness, and help football players optimize their movements during the mobile support. *e performance advantage of fog com- sports process and improve the hit rate of kicks. *rough puting [10] can just make up for the deﬁciencies of cloud this system, it can help analyze football players to quickly computing and provide real-time computing services for understand their own advantages and problems in the large-scale and distributed terminal devices. In the smart process of sports, so as to help football players optimize their factory, the low-latency characteristics of fog computing movements during the movement, increase the hit rate of are used to realize rapid analysis of manufacturing data, kicks, and ﬁnally increase the ability by 20%. timely determine the operating status of processing equipment, and provide real-time maintenance for the equipment, thus avoiding the production line shutdown 2. Evaluation Model of Football Player Training due to equipment failure. and Teaching Action Based on Artificial Intelligence 2.1.2. System Model. Based on the design idea of micro- 2.1. Artiﬁcial Intelligence services, the application is modeled as a set of services, and the calling relationship between services is expressed in the 2.1.1. Fog Calculation. Fog computing is a decentralized form of a directed acyclic graph. computing infrastructure that can utilize one or more IoT devices close to the user edge to collaboratively perform a P P x x c � 􏽘 c ∧k|D ∈ d(k). A k (1) large number of communications, control, storage, and management tasks. Data transmission has extremely low International Transactions on Electrical Energy Systems 3 In the case of considering multiple resources, it is im- possible to guarantee the maximum utilization of n types of Fog layer resources. When calculating the fogging resources, the re- source consumption of the fogging equipment is as follows: m o R � 􏽘 R ∀A |A ∈ F. fog A i i i (4) Based on the above analysis, the system average het- Clouds erogeneous resource utilization η is deﬁned as the ave weighted average of all types of resource utilization [12]. *e solution goal of the service placement strategy is to maxi- mize the average resource utilization of the system: k x Maxη � 􏽘 A · η . ave i (5) Regarding the placement of microservices in the fog User layer layer, if only the service quality indicators are considered in the modeling process, the resources on the fog layer Figure 1: Fog computing system architecture. equipment cannot be fully utilized, resulting in a decrease in the resource utilization of the fog layer equipment. *ere- fore, the task allocation problem under resource constraints is more in line with the fog computing environment. Low energy consumption 2.1.3. Artiﬁcial Intelligence Builds a Network Model. For smart factories, information technology and smart Location and emotion manufacturing equipment are the keys to smart Heterogeneity perception manufacturing. *e implementation of smart factories needs Fog computing to consider the current status quo, and production tech- nology requirements, based on the current status quo, Mobility Geographical proposed a cloud manufacturing system-oriented smart support distribution manufacturing architecture [13, 14] and established a multi- agent manufacturing system prototype platform. *e cloud Slow delay computing-based manufacturing system architecture can realize the service-oriented manufacturing system function. Figure 2: Features of fog computing. Under the combined action of demand-driven and dis- turbing factors, the cloud processing, analysis, and decision- making ensure resource utilization and system ﬂexibility. In formula (1), λ represents the ith device, and λ *e seamless integration of fog computing and cloud represents the kth user’s request rate for the service, where d computing provides an improved cloud-fog architecture (k) is the set of devices on the path from the accessed device solution. *e architecture includes four levels, from bottom to the cloud layer by user k. to top, the terminal layer, the network layer, the platform For computing resources of type R, the resource con- layer, and the application layer. At the terminal layer, the number of devices is very large, sumption on the device λ can be expressed as T , and the A ,x cap and the types are also very diverse, with very strong het- total amount of resources on the device λ is W , and the k A ,x x erogeneity. To accelerate the realization of manufacturing utilization rate of type R resources on the λ device η is k A through artiﬁcial intelligence, modular manufacturing units deﬁned as follows: are introduced into the smart factory. Today, by improving performance, the network layer, platform layer, and appli- A ,x (2) η � . cap i cation layer are processing the collected data. *e applica- A ,x tion layer processes the collected relevant data. After the application layer collects the football player data, after the If the device set of the cloud layer in the fog computing is calculation of the fog calculation, the integration and in- G and the number of devices is <F>, then the resource teraction of various information are realized through the utilization rate η of the system is the arithmetic average of technical integration of multiple networks. *e technology the utilization rates of all devices [11]: integration of multiple networks realizes the integration and 􏽐 η x A i i interaction of various information. *e fog computing of (3) η � . ⟨F⟩ 4 International Transactions on Electrical Energy Systems i e short-distance and distributed deployment not only inherits AMD(i) � 􏽘 􏽘 d . (7) cloud computing services but also has the advantages of low 􏼁 1 i�1 latency, low energy consumption, mobility, and location awareness [15]. From this, we normalize the obtained ADM (admin, refers to the super administrator) for normalization: 2.2. Football Player Training Management System. Football is a ball game in which the feet dominate the ball. AMD(i) − AMD min AMD(i) � . (8) tol Football is highly antagonistic, and athletes use various AMD − AMD max min actions permitted by the rules in the game, including running and rushing, which are equivalent to rivals for ﬁerce Among them, AMD (i) represents the cumulative competition: size of S at the ith frame, ADM(i) is the cumulative tol size after normalization, ADM is the minimum max (1) Football training should be based on the physical value in the cumulative sequence, and ADM is the min ﬁtness, basic conditions, and environmental fac- minimum value in the cumulative sequence. tors of diﬀerent players to develop relevant In order to be able to select a suitable frame, a training plans. In addition to some basic training suitable threshold value δ needs to be selected. *e items about football skills and physical ﬁtness threshold value δ is a constant between 0 and 1, and training, training should also be carried out then, the following formula can be used to determine according to diﬀerent players' positions and per- whether it is a suitable frame: sonal abilities. After the introduction of artiﬁcial intelligence-related technologies, better calcula- ADM(i) − ADM(i − 1) ≥ δ. (9) tions and manual modiﬁcations can be made to tol tol develop a training plan that is most suitable for each player’s stage. An example of the manage- *rough the formula, the frame larger than δ can be used as a suitable frame so that the remaining frames ment of the training of diﬀerent football players is shown in Figure 3. can be removed. *e eﬀect can be identiﬁed according to the size of the threshold in the ﬁgure, as Physical ﬁtness trainers can help players get better shown in Figure 4. physical ﬁtness and deal with more emergencies or As can be seen in Figure 4, with the gradual increase diﬀerent game conditions on the court. Players in the threshold, the ﬂuctuation of the recognition playing diﬀerent positions on the court require diﬀerent physical ﬁtness patterns. In addition to the rate is not very large. It can be seen that changes in the size of the threshold [17] will have a small mi- position factor, the physical ﬁtness of a player also determines which aspect of physical training he croscopic impact, and the recognition rate is the best when the threshold is around 0.16. *erefore, the needs. threshold can be controlled at 0.16 during the ex- *e purpose of frame selection [16] is to extract periment, so that the experimental data will be more frames that contain a large amount of motion in- rigorous. formation to improve the recognition eﬀect. (2) Frame Selection Model According to the sequence structure of the action, the frame with a small amount of action information An action is composed of many consecutive frames, is extracted at the beginning and end of the action. but each frame is not equally important due to delays *en, the formula representing the amount of and other reasons. *erefore, redundant frames in change from frame to frame is as follows: each action sequence need to be removed. A frame e selection model is proposed for the selection of S � 􏽘 d . (6) 􏼁 suitable frames. i�1 For the detection of diﬀerent segmentation points of continuous actions, the detection is performed In this formula, d represents the angular distance according to the sliding window and then described between frames, and S refers to the total angular by a scoring system. *en, the score of an action change between the current frame and the next sequence n in the ith type of action is as follows: frame. Sco (n) � K n, c . (10) For the same person, there will be various diﬀerences i i between the same actions, so the diﬀerence in the size According to the above formula (10), the continuous of diﬀerent samples cannot be directly used in the action sequence data of the test are extracted into the algorithm. To compare the angular distances in scoring system according to the sequence fragments diﬀerent samples on a uniform scale, and to express extracted by the sliding window, and the score size of the principle of the frame selection algorithm more each category corresponding to each frame can be vividly, the normalized representation of the accu- obtained. mulated data of various motions is as follows: Recognition rate International Transactions on Electrical Energy Systems 5 Special training management Basic training management Athlete Management Training goal Training plan setting management Training record Training record Training evaluation Player management Game Team Soccer player management management Training Training evaluation plan Pre-match training Training implementation Competition Summary after data statistics the game Figure 3: Football player training management system. After making a decision, to optimize the system, the 1.5 frame node is not allowed to oﬄoad tasks to perform 1.3 collaborative processing. *en, the condition that the 1.1 ratio of the data sharing capacity of each trans- mission path of the slave node is summed to 1 will 0.9 not exist or be meaningless. Arranging the above two 0.7 situations can be expressed as follows: 0.5 i i i i 0.3 ζ λ � λ , ∀i, n ∈ i, n|K � 1 , m ∈ M. 􏽘 􏽮 􏽯 min mnk mnk n (13) k∈k 0.1 (3) Football Action Analysis Model 0.08 0.12 0.16 0.20 Angle distance threshold In this paper, an improved OpenPose network model is used to analyze the body movements of athletes Figure 4: Recognition rate varies with the size of the threshold. during kicking with both feet [20]. *e network model is a ﬂexible way that the database model is When nodes in diﬀerent frames receive diﬀerent conceived as representing objects and their rela- types of tasks from outside the system, the sum of the tionships. Its characteristic is that it can be viewed as proportion of oﬄoading to other nodes that allow a graph whose object type is node and relationship assistance and the proportion of local processing is 1 type is arc, and it is not limited to a hierarchical [18, 19]. It can be expressed as follows: structure. *en, combined with OpenPose, you can i observe the changes in the posture of the human 􏽘 μ � 1. mon (11) body in real time. *en, the data are preprocessed to i∈ i|K �1e { } obtain the coordinate data of the key points of the limbs in the process of robust rope skipping. Because In the formula, because some nodes do not allow the coordinates of the key points of the body certain tasks to be processed, when diﬀerent frames movements are a time series and have a certain m are allowed to be processed in cooperation with connection with each other, this study ﬁnally applies the transmission type, they can be transmitted to the the ALSTM-LSTM model to the analysis of the body nodes through various paths. *e sum of the pro- movements through the algorithm transformation portions of all path allocations is 1, so it can be method in the multi-label classiﬁcation algorithm. expressed as follows: *e ALSTM-LSTM (time-series forecasting method) i i model is applied to the analysis of body movements, 􏽘 λ � 1, ∀j, n ∈ 􏽮i, n|K � 1􏽯. mnk n (12) as shown in Figure 5. k∈k 6 International Transactions on Electrical Energy Systems Start Football data stream Improved Openpose network framework Test Excellent Improve the model No network performance evaluation Yes Human body key point coordinate data Neural network framework Test Excellent model Improve the No performance network evaluation Yes Limb movement analysis Algorithm transformation method Finish Figure 5: Flow chart of football action analysis model design. 2.3. Football Player Training and Teaching Actions Based on (2) Deep Belief Network Artiﬁcial Intelligence DBN algorithm is a kind of neural network of (1) With the help of artiﬁcial intelligence for football machine learning. It can be used for unsupervised players’ training and teaching actions, detailed plan learning and supervised learning. DBN is a design and implementation can be carried out. probabilistic generative model. Compared with the Under this technology, it can be simply divided into traditional neural network of discriminant model, the following modules [21]. First, the relevant data of the generative model is to establish a relationship football players are entered into the server, and the between the observation data and the label. For the topological structure is managed according to the input feature data, usually the data are ﬁtted with a relevant operation of the server, as shown in Figure 6. presumption of what distribution it conforms to, and then, it is trained and solved according to its As shown in Figure 6, the server is used as a node to assumed distribution model. Many distributions manage diﬀerent related data of athletes, and the can take advantage of the unique properties and logical processing server of the system structure is learning process of the energy model, limiting the expanded into athlete management, athlete-related Boltzmann machine [22, 23] to be transformed data, trainer management, and background man- from the energy model. *en, the energy deﬁnition agement and is timely feedback based on diﬀerent is as follows: changes each time to improve the athlete’s training. International Transactions on Electrical Energy Systems 7 obtain from the image. *e global information of the data in Sports related the image is extracted, and a global feature extraction model data Manager [25, 26]isestablished. In action recognition, the position management of changes in key parts of the human body in the action se- course training quence are generally used for estimation. *e speciﬁc Training plan management training management system for the human body is shown in Figure 7. 3. Experiment and Analysis Backstage Athlete management Management 3.1. Data Analysis of Artiﬁcial Intelligence. Although Serverice movement is an external manifestation of the ability to collaborate, the completion of technical movement is the Figure 6: Schematic diagram of the topological structure of result of the coordinated work of nerves, muscles, and football player training management. sensations. According to the related theory of cooperation ability, the football player’s physical cooperation ability is deﬁned. In the understanding of artiﬁcial intelligence, it n m n m mainly relies on receipt data for timely feedback. For the E(k|θ) � 􏽘 a v − 􏽘 b h − 􏽘 􏽘 v h . (14) i i j j i j understanding of our neural network, we must also rely on i�1 j�1 i�1 j�1 the calculation of the network and the establishment of the model. In-depth understanding of various data can be ad- Among them, θ � [a , b ] is the parameter model of the i j justed in various physical indicators [27]. Table 1 shows the following formula: impact of diﬀerent groups of people on the impact of the − E(h|θ) introduction of artiﬁcial intelligence technology in football. (15) K(h|θ) � . From the data in Table 1, we can see whether the impact P(θ) on football players after the introduction of artiﬁcial in- Among them, P(θ) refers to the normalization factor in telligence will increase the interest of the players, and most the calculation of joint probability: importantly, it will bring about huge changes in improving − E(h|θ) the training quality of the players. *rough the investi- P(θ) � 􏽘 e . (16) gation of a large number of samples, it is found that there h�1 will be an 85% impact on improving the quality of training. *e establishment of the energy function [24] is to obtain Of course, there are other impacts similar to the use of the distribution of the input data, the input data can see the artiﬁcial intelligence technology to record relevant data of state of the layer unit, and the likelihood function is solved athletes and then provide better feedback based on the through speciﬁc calculations. *e formula can be expressed content obtained. After receiving the relevant feedback, the as follows: athlete can adjust the relevant training plan in time. *en, after receiving the relevant feedback, the athlete can adjust − E(v,h|θ) 􏽐 e (17) the relevant training plan in time. In addition to the rel- P(v, h|θ) � . Z(θ) evant changes in the training, the athlete can adjust the action in time during the training to reduce the physical Knowing the input data of the RBM, you can see the state injury. of the layer unit. Because of the structural characteristics of no connection in the layer and full connection between the layers, the formula for obtaining the state of the hidden layer 3.2. Training Behavior of Football Players. After the intro- unit is as follows: duction of artiﬁcial intelligence, athletes will reduce their usual mistakes, etc. *e impact of improving on the basis of ⎣ ⎦ ⎡ ⎤ P v � 1|v, θ􏼁 � μ a + 􏽘 v . (18) i i i existing levels by trainers is shown in Table 2. According to the data in the table, the inﬂuence of the research objects of diﬀerent competition levels can be ob- In the formula, μ is the activation function. According to served. First of all, certain research is conducted on whether the state of the hidden layer unit, the formula for obtaining the changes brought about by the introduction of artiﬁcial the visible layer unit in the reverse direction is as follows: intelligence will have an impact. Experimental data show that after the introduction of artiﬁcial intelligence, the ⎢ ⎥ ⎡ ⎢ ⎤ ⎥ ⎣ ⎦ P v � 1|h, θ􏼁 � μ b + 􏽘 h . (19) i j j changes brought about by both professional leagues and campus leagues are very large. For example, for school In motion recognition, motion data are a kind of real- leagues, the main impact is about 30%. In addition, the time data. *e image-based representation method means changes brought by the training team members introducing that the original data come from the image, and the relevant artiﬁcial intelligence technology for training are shown in model is established to extract the information we want to Table 3. 8 International Transactions on Electrical Energy Systems Training record Physical Trainer Training trainer Physical Adjust the Training Formulate plan Record Manage Player Belong Figure 7: Football player training management system database. Table 1: Surveyed football players on the impact of the introduction of artiﬁcial intelligence technology. Percentage Projects Number of samples (%) Increase interest in training 1008 83 Improve training quality 2013 85 Reduce injuries 1111 82 More comfortable to wear 1018 25 Complete training plan better 1339 52 Table 2: Level of the football players surveyed and the impact of the introduction of artiﬁcial intelligence. Level of inﬂuence Business league School league College league Professional league Percentage Very small 1600 1220 3311 0 10% Smaller 0 1302 4025 3000 25% Bigger 2221 1400 2030 2845 75% Very big 2012 1502 2050 3102 0 Percentage 42% 30% 10% 18% 100% Table 3: About the survey results after the use of artiﬁcial intelligence equipment in auxiliary training competitions in my country. Select item Number of surveys Survey percentage (%) Better reﬂect the problems that arise in the game 500 100 Improve players’ motivation for training 300 45 Make better training plans 130 72 Improve training eﬃciency 300 20 Reduce athletes’ injuries 90 15 Table 4: Athlete detection statistics after the adversarial personal improve the training eﬃciency of football players and ball control test experiment. improve the enthusiasm of the players in training. In addition to the intensity of the original training, it can MSQ Test group Control group Sample T value also reduce the chance of athletes being injured. K1 8.26± 1.13 8.45± 3.21 0.982 0.028 *rough the adjustment of big data, certain feedback K2 9.32± 2.21 8.23± 0.88 1.103 0.037 from athletes can be obtained. In turn, the formation of a K3 7.28± 2.28 6.63± 1.48 0.656 0.037 good training closed-loop promotion can not only improve the training eﬃciency of the athletes but also According to Table 3, it can be seen that artiﬁcial help the athletes to increase their interest in many as- intelligence equipment has a great auxiliary eﬀect on pects [28]. football. After using artiﬁcial intelligence, based on the *rough the above table method, the introduction of related training of the original football players, it can artiﬁcial intelligence in football can bring about very big International Transactions on Electrical Energy Systems 9 11 11 11 11 11 11 11 10 10 10 10 10 10 10 10 10 10 9 9 99 9 9 9 9 9 9 8 8 8 8 88 8 10 8 8 8 8 8 8 7 7 7 7 7 66 66 666 6 123456789 10 11 12 123456789 10 11 12 Test group (a) Test group (c) Test group (b) Test group (d) Control group Control group Figure 8: Comparison of capacity changes before and after getting rid of the connection test. Sliding size 5 101520253035 Accuracy Score (1) Score (2) Figure 9: Selection of diﬀerent frame numbers for sliding window. changes. It is not only reﬂected in the behavioral teaching of experimental group, the lowest number of successes is 7 football players’ training but also can better stimulate the times, and the lowest number of successes in the control potential of the players when playing games and improve the group is 6 times. *e comparison shows that the number of physical performance to a new height. At the same time, it is successes of the experimental group is signiﬁcantly better than that of the control group. *erefore, according to the not only reﬂected in sports competitions but also in pro- tecting the athletes’ bodies to ensure that they will not be results of this experiment, it can be seen that the intro- harmed. *is timely feedback can help trainers and athletes duction of artiﬁcial intelligence technology can improve the to collaborate better. athlete’s antagonistic personal ball control ability. From the perspective of diﬀerent experimental groups and control groups, there is a gap in the number of athletes’ 3.3. Experimental Football Players’ Ability Test Results after confrontational escape transfers. Observing the data in the Artiﬁcial Intelligence Training. After the experiment, the chart, we can see that the data of the experimental group are analysis of the test results of football players’ personal ball more stable, and at the same time, it is better than the control is to detect the antagonistic personal ball control of experimental group, as shown in Figure 8. diﬀerent groups of players using a paired sample test. K1 is Because football is a long-term sequence analysis [29] the preexperimental test result of the number of successful process, it is necessary to use a sliding window to segment transfers after the athlete gets rid of the test, K2 is the test diﬀerent data. *erefore, to ﬁnd the appropriate length of result of the number of athletes successfully shed the ball in the sliding window, the accumulated coordinates in diﬀerent the test before the test, and K3 is the test result of the athlete’s frames are obtained, as shown in Figure 9. catch-and-stop quality before the test in the test, as shown in Table 4. 4. Discussion From Table 4, it is observed that in the test about athletes’ antagonistic personal ball control, the number of successes *is article is dedicated to introducing artiﬁcial intelligence after the athletes get rid of the number of transfer experi- technology into the teaching behavior of football sports ments is basically between 10. According to the data in the training. *e research in the article shows that when 10 International Transactions on Electrical Energy Systems [6] M. Seyedmahmoudian, B. Horan, T. K. Soon et al., “State of computer technology and football are combined, certain the art artiﬁcial intelligence-based MPPT techniques for advantages can be obtained in the individual physical in- mitigating partial shading eﬀects on PV systems—a review,” dicators of athletes, and better feedback can be obtained Renewable and Sustainable Energy Reviews, vol. 64, pp. 435– during training to form a closed-loop virtuous circle. 455, 2016. *rough the analysis of diﬀerent frame numbers, the ath- [7] M. Wang and Z. Liang, “An extraction method of volleyball lete’s range of activities is captured, and then, quantitative spiking trajectory and teaching based on wireless sensor analysis is carried out, so that football training can get rid of network,” Security and Communication Networks, vol. 2021, purely relying on experience and enter the era of theory and no. 9, Article ID 9966994, 11 pages, 2021. digitalization. [8] V. Patel and V. Savsani, “Multi-objective optimization of a Stirling heat engine using TS-TLBO (tutorial training and self learning inspired teaching-learning based optimization) al- 5. Conclusions gorithm,” Energy, vol. 95, pp. 528–541, 2016. *e data acquired by each device can simulate the entire [9] H. Lu, Y. Li, M. Chen, H. Kim, and S. Serikawa, “Brain in- process of the athlete from before touching the ball to after telligence: go beyond artiﬁcial intelligence,” Mobile Networks and Applications, vol. 23, no. 2, pp. 368–375, 2017. ejection. *e height and distance of the virtual football ﬂight [10] X. Liang, L. Haiping, J. Liu, and L. Lin, “Reform of English are simulated according to the football players’ exertion habits interactive teaching mode based on cloud computing artiﬁcial and contact points during sports. Artiﬁcial intelligence intelligence—a practice analysis,” Journal of Intelligent and technology actually simulates the physical characteristics of Fuzzy Systems, vol. 40, no. 2, pp. 3617–3629, 2021. football players’ height, body shape, etc., according to their [11] B. Li, Y. Fei, and H. Liu, “An artiﬁcial intelligence based model joints and their spatial positions, and appears in the simulated for evaluation of college students’ ability and characteristics environment at a speciﬁc speed. *rough the introduction of through teaching evaluation,” Journal of Intelligent and Fuzzy artiﬁcial intelligence technology, the use of multi-person Systems, vol. 40, no. 2, pp. 3397–3407, 2021. whole-body positioning and interaction systems, and online [12] Z. Chen, Y. Lian, and Z. Lin, “Research on business English training, you can see the actions of other operators and load autonomous learning based on artiﬁcial intelligence and and update the training scene in real time. Football training or improved BP network model,” Journal of Intelligent and Fuzzy games can be carried out online, which can help users visually Systems, vol. 40, no. 3, pp. 1–12, 2020. analyze the on-site situation and make correct decisions about [13] L. Yu and N. Peng, “Research on English teaching reform based on artiﬁcial intelligence matching model,” Journal of passing height, angle, and time. Intelligent and Fuzzy Systems, no. 1, pp. 1–10, 2021. [14] G. Yunfeng and L. Jing, “Recognition of teaching method Data Availability eﬀects based on grid model simpliﬁcation and artiﬁcial in- telligence,” Journal of Intelligent and Fuzzy Systems, vol. 40, *e data that support the ﬁndings of this study are available no. 4, pp. 6707–6717, 2021. from the corresponding author upon reasonable request. [15] Z. Zhang, “Research and application of capability evaluation model based on BP neural network educational tech- nology—taking the problem-based learning teaching process Conflicts of Interest learning as an example,” Journal of Computational and *e authors declare that there are no conﬂicts of interest <eoretical Nanoscience, vol. 13, no. 9, pp. 6210–6217, 2016. with respect to the research, authorship, and/or publication [16] S. Zou, “Designing and practice of a college English teaching platform based on artiﬁcial intelligence,” Journal of Compu- of this article. tational and <eoretical Nanoscience, vol. 14, no. 1, pp. 104– 108, 2017. References [17] Y. Ding, “Performance analysis of public management teaching practice training based on artiﬁcial intelligence [1] D. Hassabis, D. Kumaran, C. Summerﬁeld, and M. Botvinick, technology,” Journal of Intelligent and Fuzzy Systems, vol. 40, “Neuroscience-inspired artiﬁcial intelligence,” Neuron, no. 2, pp. 3787–3800, 2021. vol. 95, no. 2, pp. 245–258, 2017. [18] J. Zhang and K. B. Letaief, “Mobile edge intelligence and [2] L. D. Raedt, K. Kersting, S. Natarajan, and D. Poole, “Sta- computing for the internet of vehicles,” Proceedings of the tistical relational artiﬁcial intelligence: logic, probability, and IEEE, vol. 99, pp. 1–16, 2019. computation,” Synthesis Lectures on Artiﬁcial Intelligence and [19] N. Dhieb, H. Ghazzai, H. Besbes, and Y. Massoud, “A secure Machine Learning, vol. 10, no. 2, pp. 1–189, 2016. AI-driven architecture for automated insurance systems: [3] L. Rongpeng, Z. Zhao, X. Zhou et al., “Intelligent 5G: when fraud detection and risk measurement,” IEEE Access, vol. 8, cellular networks meet artiﬁcial intelligence,” IEEE Wireless pp. 58546–58558, 2020. Communications, vol. 24, no. 5, pp. 175–183, 2017. [20] X. Qiao, Y. Huang, S. Dustdar, and J. Chen, “6G vision: an AI- [4] J. H. *rall, X. Li, Q. Li et al., “Artiﬁcial intelligence and driven decentralized network and service architecture,” IEEE machine learning in radiology: opportunities, challenges, Internet Computing, vol. 24, no. 4, pp. 33–40, 2020. pitfalls, and criteria for success,” Journal of the American [21] D. Y. Chao, “Computation of elementary siphons in petri nets College of Radiology, vol. 15, no. 3, pp. 504–508, 2018. for deadlock control,” <e Computer Journal, vol. 49, no. 4, [5] P. Glauner, J. A. Meira, P. Valtchev, R. State, and F. Bettinger, “*e challenge of non-technical loss detection using artiﬁcial pp. 470–479, 2006. [22] S. M. Senouci, H. Sedjelmaci, J. Liu, M. H. Rehmani, and intelligence: a survey,” International Journal of Computational Intelligence Systems, vol. 10, no. 1, pp. 760–775, 2017. E. Bou-Harb, “AI-driven cybersecurity threats to future International Transactions on Electrical Energy Systems 11 networks [from the guest editors],” IEEE Vehicular Tech- nology Magazine, vol. 15, no. 3, pp. 5-6, 2020. [23] K. Karras, E. Pallis, G. Mastorakis et al., “A hardware ac- celeration platform for AI-based inference at the edge,” Circuits, Systems, and Signal Processing, vol. 39, no. 2, pp. 1059–1070, 2020. [24] Y. Zhou and X. Zhu, “Analysis of vehicle network architecture and performance optimization based on soft deﬁnition of integration of cloud and fog,” IEEE Access, vol. 7, pp. 101171–101177, 2019. [25] D. Zeng, L. Gu, S. Guo, Z. Cheng, and S. Yu, “Joint opti- mization of task scheduling and image placement in fog computing supported software-deﬁned embedded system,” IEEE Transactions on Computers, vol. 65, no. 12, pp. 3702– 3712, 2016. [26] Y. Ramzanpoor, M. Hosseini Shirvani, and M. Golsorkhtabaramiri, “Multi-objective fault-tolerant opti- mization algorithm for deployment of IoT applications on fog computing infrastructure,” Complex & Intelligent Systems, vol. 8, no. 1, pp. 361–392, 2021. [27] M. K. Hussein and M. H. Mousa, “Eﬃcient task oﬄoading for IoT-based applications in fog computing using ant colony optimization,” IEEE Access, vol. 8, pp. 37191–37201, 2020. [28] C. Guerrero, I. Lera, and C. Juiz, “A lightweight decentralized service placement policy for performance optimization in fog computing,” Journal of Ambient Intelligence and Humanized Computing, vol. 10, no. 6, pp. 2435–2452, 2019. [29] T. Huang, W. Lin, C. Xiong, R. Pan, and J. Huang, “An ant colony optimization-based multiobjective service replicas placement strategy for fog computing,” IEEE Transactions on Cybernetics, vol. 51, pp. 1–14, 2020. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Transactions on Electrical Energy Systems Hindawi Publishing Corporation http://www.deepdyve.com/lp/hindawi-publishing-corporation/evaluation-model-of-football-players-training-and-teaching-actions-i90dAWx8Hz

Loading next page...

References (30)

Yunfeng Guo, Jing Li (2021)
Recognition of teaching method effects based on grid model simplification and artificial intelligence
J. Intell. Fuzzy Syst., 40
Arti ﬁ cial Intelligence and Machine Learning in Radiology: Opportunities, Challenges, Pitfalls, and Criteria for Success
K. Kersting L. D. Raedt (2016)
Statistical relational artificial intelligence: logic, probability, and computation,
Synthesis Lectures on Artificial Intelligence and Machine Learning, 10
Rongpeng Li, Zhifeng Zhao, Xuan Zhou, Guoru Ding, Yan Chen, Zhongyao Wang, Honggang Zhang (2017)
Intelligent 5G: When Cellular Networks Meet Artificial Intelligence
IEEE Wireless Communications, 24
Yaser Ramzanpoor, Mirsaeid Shirvani, M. Golsorkhtabaramiri (2021)
Multi-objective fault-tolerant optimization algorithm for deployment of IoT applications on fog computing infrastructure
Complex & Intelligent Systems, 8
Lan Yu, Ning Peng (2021)
Research on English teaching reform based on artificial intelligence matching model
Journal of Intelligent and Fuzzy Systems
D. Hassabis, D. Kumaran, C. Summerfield, M. Botvinick (2017)
Neuroscience-Inspired Artificial Intelligence
Neuron, 95
Shuli Zou (2017)
Designing and Practice of a College English Teaching Platform Based on Artificial Intelligence
Journal of Computational and Theoretical Nanoscience, 14
Xiuquan Qiao, Yakun Huang, S. Dustdar, Junliang Chen (2020)
6G Vision: An AI-Driven Decentralized Network and Service Architecture
IEEE Internet Computing, 24
Xiufang Liang, Haiping Lv, Jie Liu, Lin Lin (2021)
Reform of English interactive teaching mode based on cloud computing artificial intelligence - a practice analysis
J. Intell. Fuzzy Syst., 40
Tiansheng Huang, Weiwei Lin, Chen Xiong, Rui Pan, Jingxuan Huang (2020)
An Ant Colony Optimization-Based Multiobjective Service Replicas Placement Strategy for Fog Computing
IEEE Transactions on Cybernetics, 51
Patrick Glauner, Andre Boechat, Lautaro Dolberg, J. Meira, R. State, Franck Bettinger, Yves Rangoni, Diogo Duarte (2016)
The Challenge of Non-Technical Loss Detection using Artificial Intelligence: A Survey
Int. J. Comput. Intell. Syst., 10
Huimin Lu, Yujie Li, Min Chen, Hyoungseop Kim, S. Serikawa (2017)
Brain Intelligence: Go beyond Artificial Intelligence
Mobile Networks and Applications, 23
L. Raedt, K. Kersting, Sriraam Natarajan, D. Poole (2016)
Statistical Relational Artificial Intelligence: Logic, Probability, and Computation
Meifang Wang, Zhange Liang (2021)
An Extraction Method of Volleyball Spiking Trajectory and Teaching Based on Wireless Sensor Network
Secur. Commun. Networks, 2021
Jun Zhang, K. Letaief (2019)
Mobile Edge Intelligence and Computing for the Internet of Vehicles
Proceedings of the IEEE, 108
Zhongmei Zhang (2016)
Research and Application of Capability Evaluation Model Based on BP Neural Network Educational Technology—Taking the Problem-Based Learning Teaching Process Learning as an Example
Journal of Computational and Theoretical Nanoscience, 13
Najmeddine Dhieb, Hakim Ghazzai, Hichem Besbes, Y. Massoud (2020)
A Secure AI-Driven Architecture for Automated Insurance Systems: Fraud Detection and Risk Measurement
IEEE Access, 8
Yi Zhou, Xianglei Zhu (2019)
Analysis of Vehicle Network Architecture and Performance Optimization Based on Soft Definition of Integration of Cloud and Fog
IEEE Access, 7
Kimon Karras, E. Pallis, G. Mastorakis, Y. Nikoloudakis, J. Batalla, C. Mavromoustakis, E. Markakis (2020)
A Hardware Acceleration Platform for AI-Based Inference at the Edge
Circuits, Systems, and Signal Processing, 39
M. Seyedmahmoudian, B. Horan, T. Soon, R. Rahmani, A. Oo, S. Mekhilef, A. Stojcevski (2016)
State of the art artificial intelligence-based MPPT techniques for mitigating partial shading effects on PV systems – A review
Renewable & Sustainable Energy Reviews, 64
Deze Zeng, Lin Gu, Song Guo, Zixue Cheng, Shui Yu (2016)
Joint Optimization of Task Scheduling and Image Placement in Fog Computing Supported Software-Defined Embedded System
IEEE Transactions on Computers, 65
Bin Li, Yan Fei, Hui Liu (2021)
An artificial intelligence based model for evaluation of college students' ability and characteristics through teaching evaluation
J. Intell. Fuzzy Syst., 40
Vivek Patel, V. Savsani (2016)
Multi-objective optimization of a Stirling heat engine using TS-TLBO (tutorial training and self learning inspired teaching-learning based optimization) algorithm
Energy, 95
M. Hussein, M. Mousa (2020)
Efficient Task Offloading for IoT-Based Applications in Fog Computing Using Ant Colony Optimization
IEEE Access, 8
S. Senouci, Hichem Sedjelmaci, Jiajia Liu, M. Rehmani, E. Bou-Harb (2020)
AI-Driven Cybersecurity Threats to Future Networks [From the Guest Editors]
IEEE Vehicular Technology Magazine
D. Chao (2006)
Computation of Elementary Siphons in Petri Nets For Deadlock Control
Comput. J., 49
Carlos Guerrero, I. Lera, C. Juiz (2018)
A lightweight decentralized service placement policy for performance optimization in fog computing
Journal of Ambient Intelligence and Humanized Computing, 10
Zijuan Chen, Ying Lian, Zhipeng Lin (2021)
Research on business English autonomous learning based on artificial intelligence and improved BP network model
J. Intell. Fuzzy Syst., 40
Yahong Ding (2021)
Performance analysis of public management teaching practice training based on artificial intelligence technology
J. Intell. Fuzzy Syst., 40

Publisher: Hindawi Publishing Corporation
Copyright: Copyright © 2022 Yun Feng and Yongan Wang. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
eISSN: 2050-7038
DOI: 10.1155/2022/7427967
Publisher site: See Article on Publisher Site

Abstract

Hindawi International Transactions on Electrical Energy Systems Volume 2022, Article ID 7427967, 11 pages https://doi.org/10.1155/2022/7427967 Research Article Evaluation Model of Football Players’ Training and Teaching Actions Based on Artificial Intelligence Yun Feng and Yongan Wang College of Physical Education, China West Normal University, Nanchong 637002, Sichuan, China Correspondence should be addressed to Yun Feng; fengyun1991@cwnu.edu.cn Received 10 June 2022; Revised 9 July 2022; Accepted 18 July 2022; Published 30 August 2022 Academic Editor: Raghavan Dhanasekaran Copyright © 2022 Yun Feng and Yongan Wang. *is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Football is a sport that needs to combine the physical stamina and physical characteristics of athletes. It needs to pay attention to the diﬀerences in diﬀerent individuals and then conduct targeted training. In this regard, this article introduces artiﬁcial in- telligence technology into the teaching actions of football player training and analyzes the teaching elements according to the characteristics of the football player’s body. *rough the analysis of fog computing under artiﬁcial intelligence, this article aimed to study the related eﬀects of combining intelligent technology on the basis of athletes’ original training. *is article proposes the establishment of a system model and quantitative analysis of diﬀerent frames of football players’ movements. According to the combination of two-phase analysis, it can be concluded that after the introduction of artiﬁcial intelligence technology, the ability of football players in various indexes has increased by 20%. more eﬀective contact methods during training; on the other 1. Introduction hand, it provides more diversiﬁed training methods for With the development and application of computer tech- athletes in future training, so that players have higher in- nology, artiﬁcial intelligence technology has been widely terest and enthusiasm in training, so that athletes can be used in sports training. Under the premise of improving the more professional and precise in football performance. At competitiveness of football, it is very urgent to seek to the same time, it provides a certain theoretical basis for the improve the level of training. Observing a lot of training data sports group and opens up new research points for the study can show that the commonality of basic training needs to be of football player training and teaching. Based on the above strengthened. *erefore, in the context of the improvement related content, Hassabis et al. found that the ﬁelds of of the global football level and the diversiﬁcation of the neuroscience and AI have a long history [1]. In addition, training process, the lack of personalized training methods Raedt et al. believed that intelligent agents interacting with and content has been unable to adapt to the trend of football the real world will encounter individuals, courses, test re- sults, etc., need to reason about the attributes of these in- development. *is method will aﬀect the improvement of athletes’ abilities to a certain extent, and it is very necessary dividuals and the relationship between them, and deal with to introduce artiﬁcial intelligence technology into football uncertainty [2]. At the same time, Rongpeng et al. believed training. that 5G is a key enabler and infrastructure provider in the On the one hand, by visually expressing the training ICT industry by providing various services with diﬀerent content of the vertical and horizontal movements of football needs [3]. In addition to the abovementioned views on AI, in the form of data, the research is conducive to the im- *rall et al. believed that it is driven by the availability of provement of the training ability of football telemobilization large data sets (“big data”), signiﬁcant advances in com- under the condition of artiﬁcial intelligence. In turn, it puting power, and new deep learning algorithms. Global provides a theoretical basis for football coaches to provide interest in artiﬁcial intelligence (AI) applications, including 2 International Transactions on Electrical Energy Systems imaging, is high and growing rapidly. In addition to de- latency, and fog computing has a vast geographic distri- veloping new AI methods themselves, they are also faced bution and a large-scale sensor network with a large number with better ways to share image data and standards for of network nodes. At the same time, through the connection validating the use of AI programs across diﬀerent imaging between the fog node and the device, fog computing can platforms [4]. Since this article mainly introduces artiﬁcial reduce the processing burden of resource-constrained de- intelligence research, Glauner et al.’s main research direction vices, meet the requirements of delay-sensitive applications, is to use artiﬁcial intelligence (AI) to solve diﬀerent prob- and overcome the bandwidth limitation of centralized lems. Finally, it investigated these research works in a services. *e basic framework of fog computing is similar to comprehensive review of the algorithms, features, and data cloud computing, but its lower-level architecture has special sets used [5]. In view of the fragility of the above content, and components that are sensitive to time response. With this in view of the great interest in artiﬁcial intelligence by feature, you can further control and enhance services such as Seyedmahmoudian et al., it links football and artiﬁcial in- sports. Figure 1 shows a mode of fog calculation. telligence technology together. Extensive research has been In Figure 1, there are three layers as a whole: cloud layer, conducted on various ways of football. According to the fog layer, and user layer. Combining the fog calculation with robust, reliable, and fast performance of the artiﬁcial in- football sports can provide timely feedback of sports-related telligence-based MPPT method, the motion posture of the data, thereby forming a benign closed loop. *e cloud layer human body is discussed under various conditions [6]. is a computing cluster composed of large servers, located at Wang and Liang obtained the upper limb trajectory in 3D the top of the entire architecture. *e fog layer is mainly space and realized the upper limb trajectory extraction [7]. composed of various embedded computing devices, located Because artiﬁcial intelligence and algorithms are inevitably between the user and the cloud, mainly provides a de- inseparable, Patel and Savsani proposed a Stirling multi- ployment environment for delay-sensitive applications, and objective optimization strategy and introduced the multi- responds quickly to user requests. According to the physical objective TS-TLBO algorithm to get the real Pareto frontier. distance between the fog layer device and the user, the fog By comparing the optimization strategies between diﬀerent layer device is subdivided into multiple sublayers, and the closer to the user side, the lower the network delay of the algorithms, the relevant research methods of artiﬁcial in- telligence are proposed, and the best combination with the sublayer. *e network communication overhead between teaching design of football players is obtained and then put devices in the same sublayer is signiﬁcantly lower than the forward the relevant model establishment formula to cross-layer network communication overhead [1, 9]. *e quantify the athlete’s movement posture [8]. user layer directly interacts with the user, sends the collected *e innovative points of this article are as follows: (1) data to the upper layer, and receives the returned result after computer technology is used to establish a systematic study the upper layer processing, as shown in Figure 2. of the actions of football players. Personalized research on *e fog is located between the terminal device and the athletes with diﬀerent personal information is conducted, cloud and builds a bridge between the terminal device and collected system management information is managed, the cloud. Fog computing, like cloud computing, can player training content is recorded, and model analysis on provide computing, storage, and network services for the collected information is performed. (2) *e constructed terminal devices. Unlike cloud computing, fog computing network model is used to design and develop diﬀerent is closer to terminal equipment. *e main characteristics of teaching systems. *rough this system, it can help the fog computing are shown in the ﬁgure: short-distance, analysis of football players to quickly understand their own distributed deployment makes fog computing have many advantages and problems in the process of sports, so as to advantages, including low latency, location awareness, and help football players optimize their movements during the mobile support. *e performance advantage of fog com- sports process and improve the hit rate of kicks. *rough puting [10] can just make up for the deﬁciencies of cloud this system, it can help analyze football players to quickly computing and provide real-time computing services for understand their own advantages and problems in the large-scale and distributed terminal devices. In the smart process of sports, so as to help football players optimize their factory, the low-latency characteristics of fog computing movements during the movement, increase the hit rate of are used to realize rapid analysis of manufacturing data, kicks, and ﬁnally increase the ability by 20%. timely determine the operating status of processing equipment, and provide real-time maintenance for the equipment, thus avoiding the production line shutdown 2. Evaluation Model of Football Player Training due to equipment failure. and Teaching Action Based on Artificial Intelligence 2.1.2. System Model. Based on the design idea of micro- 2.1. Artiﬁcial Intelligence services, the application is modeled as a set of services, and the calling relationship between services is expressed in the 2.1.1. Fog Calculation. Fog computing is a decentralized form of a directed acyclic graph. computing infrastructure that can utilize one or more IoT devices close to the user edge to collaboratively perform a P P x x c � 􏽘 c ∧k|D ∈ d(k). A k (1) large number of communications, control, storage, and management tasks. Data transmission has extremely low International Transactions on Electrical Energy Systems 3 In the case of considering multiple resources, it is im- possible to guarantee the maximum utilization of n types of Fog layer resources. When calculating the fogging resources, the re- source consumption of the fogging equipment is as follows: m o R � 􏽘 R ∀A |A ∈ F. fog A i i i (4) Based on the above analysis, the system average het- Clouds erogeneous resource utilization η is deﬁned as the ave weighted average of all types of resource utilization [12]. *e solution goal of the service placement strategy is to maxi- mize the average resource utilization of the system: k x Maxη � 􏽘 A · η . ave i (5) Regarding the placement of microservices in the fog User layer layer, if only the service quality indicators are considered in the modeling process, the resources on the fog layer Figure 1: Fog computing system architecture. equipment cannot be fully utilized, resulting in a decrease in the resource utilization of the fog layer equipment. *ere- fore, the task allocation problem under resource constraints is more in line with the fog computing environment. Low energy consumption 2.1.3. Artiﬁcial Intelligence Builds a Network Model. For smart factories, information technology and smart Location and emotion manufacturing equipment are the keys to smart Heterogeneity perception manufacturing. *e implementation of smart factories needs Fog computing to consider the current status quo, and production tech- nology requirements, based on the current status quo, Mobility Geographical proposed a cloud manufacturing system-oriented smart support distribution manufacturing architecture [13, 14] and established a multi- agent manufacturing system prototype platform. *e cloud Slow delay computing-based manufacturing system architecture can realize the service-oriented manufacturing system function. Figure 2: Features of fog computing. Under the combined action of demand-driven and dis- turbing factors, the cloud processing, analysis, and decision- making ensure resource utilization and system ﬂexibility. In formula (1), λ represents the ith device, and λ *e seamless integration of fog computing and cloud represents the kth user’s request rate for the service, where d computing provides an improved cloud-fog architecture (k) is the set of devices on the path from the accessed device solution. *e architecture includes four levels, from bottom to the cloud layer by user k. to top, the terminal layer, the network layer, the platform For computing resources of type R, the resource con- layer, and the application layer. At the terminal layer, the number of devices is very large, sumption on the device λ can be expressed as T , and the A ,x cap and the types are also very diverse, with very strong het- total amount of resources on the device λ is W , and the k A ,x x erogeneity. To accelerate the realization of manufacturing utilization rate of type R resources on the λ device η is k A through artiﬁcial intelligence, modular manufacturing units deﬁned as follows: are introduced into the smart factory. Today, by improving performance, the network layer, platform layer, and appli- A ,x (2) η � . cap i cation layer are processing the collected data. *e applica- A ,x tion layer processes the collected relevant data. After the application layer collects the football player data, after the If the device set of the cloud layer in the fog computing is calculation of the fog calculation, the integration and in- G and the number of devices is <F>, then the resource teraction of various information are realized through the utilization rate η of the system is the arithmetic average of technical integration of multiple networks. *e technology the utilization rates of all devices [11]: integration of multiple networks realizes the integration and 􏽐 η x A i i interaction of various information. *e fog computing of (3) η � . ⟨F⟩ 4 International Transactions on Electrical Energy Systems i e short-distance and distributed deployment not only inherits AMD(i) � 􏽘 􏽘 d . (7) cloud computing services but also has the advantages of low 􏼁 1 i�1 latency, low energy consumption, mobility, and location awareness [15]. From this, we normalize the obtained ADM (admin, refers to the super administrator) for normalization: 2.2. Football Player Training Management System. Football is a ball game in which the feet dominate the ball. AMD(i) − AMD min AMD(i) � . (8) tol Football is highly antagonistic, and athletes use various AMD − AMD max min actions permitted by the rules in the game, including running and rushing, which are equivalent to rivals for ﬁerce Among them, AMD (i) represents the cumulative competition: size of S at the ith frame, ADM(i) is the cumulative tol size after normalization, ADM is the minimum max (1) Football training should be based on the physical value in the cumulative sequence, and ADM is the min ﬁtness, basic conditions, and environmental fac- minimum value in the cumulative sequence. tors of diﬀerent players to develop relevant In order to be able to select a suitable frame, a training plans. In addition to some basic training suitable threshold value δ needs to be selected. *e items about football skills and physical ﬁtness threshold value δ is a constant between 0 and 1, and training, training should also be carried out then, the following formula can be used to determine according to diﬀerent players' positions and per- whether it is a suitable frame: sonal abilities. After the introduction of artiﬁcial intelligence-related technologies, better calcula- ADM(i) − ADM(i − 1) ≥ δ. (9) tions and manual modiﬁcations can be made to tol tol develop a training plan that is most suitable for each player’s stage. An example of the manage- *rough the formula, the frame larger than δ can be used as a suitable frame so that the remaining frames ment of the training of diﬀerent football players is shown in Figure 3. can be removed. *e eﬀect can be identiﬁed according to the size of the threshold in the ﬁgure, as Physical ﬁtness trainers can help players get better shown in Figure 4. physical ﬁtness and deal with more emergencies or As can be seen in Figure 4, with the gradual increase diﬀerent game conditions on the court. Players in the threshold, the ﬂuctuation of the recognition playing diﬀerent positions on the court require diﬀerent physical ﬁtness patterns. In addition to the rate is not very large. It can be seen that changes in the size of the threshold [17] will have a small mi- position factor, the physical ﬁtness of a player also determines which aspect of physical training he croscopic impact, and the recognition rate is the best when the threshold is around 0.16. *erefore, the needs. threshold can be controlled at 0.16 during the ex- *e purpose of frame selection [16] is to extract periment, so that the experimental data will be more frames that contain a large amount of motion in- rigorous. formation to improve the recognition eﬀect. (2) Frame Selection Model According to the sequence structure of the action, the frame with a small amount of action information An action is composed of many consecutive frames, is extracted at the beginning and end of the action. but each frame is not equally important due to delays *en, the formula representing the amount of and other reasons. *erefore, redundant frames in change from frame to frame is as follows: each action sequence need to be removed. A frame e selection model is proposed for the selection of S � 􏽘 d . (6) 􏼁 suitable frames. i�1 For the detection of diﬀerent segmentation points of continuous actions, the detection is performed In this formula, d represents the angular distance according to the sliding window and then described between frames, and S refers to the total angular by a scoring system. *en, the score of an action change between the current frame and the next sequence n in the ith type of action is as follows: frame. Sco (n) � K n, c . (10) For the same person, there will be various diﬀerences i i between the same actions, so the diﬀerence in the size According to the above formula (10), the continuous of diﬀerent samples cannot be directly used in the action sequence data of the test are extracted into the algorithm. To compare the angular distances in scoring system according to the sequence fragments diﬀerent samples on a uniform scale, and to express extracted by the sliding window, and the score size of the principle of the frame selection algorithm more each category corresponding to each frame can be vividly, the normalized representation of the accu- obtained. mulated data of various motions is as follows: Recognition rate International Transactions on Electrical Energy Systems 5 Special training management Basic training management Athlete Management Training goal Training plan setting management Training record Training record Training evaluation Player management Game Team Soccer player management management Training Training evaluation plan Pre-match training Training implementation Competition Summary after data statistics the game Figure 3: Football player training management system. After making a decision, to optimize the system, the 1.5 frame node is not allowed to oﬄoad tasks to perform 1.3 collaborative processing. *en, the condition that the 1.1 ratio of the data sharing capacity of each trans- mission path of the slave node is summed to 1 will 0.9 not exist or be meaningless. Arranging the above two 0.7 situations can be expressed as follows: 0.5 i i i i 0.3 ζ λ � λ , ∀i, n ∈ i, n|K � 1 , m ∈ M. 􏽘 􏽮 􏽯 min mnk mnk n (13) k∈k 0.1 (3) Football Action Analysis Model 0.08 0.12 0.16 0.20 Angle distance threshold In this paper, an improved OpenPose network model is used to analyze the body movements of athletes Figure 4: Recognition rate varies with the size of the threshold. during kicking with both feet [20]. *e network model is a ﬂexible way that the database model is When nodes in diﬀerent frames receive diﬀerent conceived as representing objects and their rela- types of tasks from outside the system, the sum of the tionships. Its characteristic is that it can be viewed as proportion of oﬄoading to other nodes that allow a graph whose object type is node and relationship assistance and the proportion of local processing is 1 type is arc, and it is not limited to a hierarchical [18, 19]. It can be expressed as follows: structure. *en, combined with OpenPose, you can i observe the changes in the posture of the human 􏽘 μ � 1. mon (11) body in real time. *en, the data are preprocessed to i∈ i|K �1e { } obtain the coordinate data of the key points of the limbs in the process of robust rope skipping. Because In the formula, because some nodes do not allow the coordinates of the key points of the body certain tasks to be processed, when diﬀerent frames movements are a time series and have a certain m are allowed to be processed in cooperation with connection with each other, this study ﬁnally applies the transmission type, they can be transmitted to the the ALSTM-LSTM model to the analysis of the body nodes through various paths. *e sum of the pro- movements through the algorithm transformation portions of all path allocations is 1, so it can be method in the multi-label classiﬁcation algorithm. expressed as follows: *e ALSTM-LSTM (time-series forecasting method) i i model is applied to the analysis of body movements, 􏽘 λ � 1, ∀j, n ∈ 􏽮i, n|K � 1􏽯. mnk n (12) as shown in Figure 5. k∈k 6 International Transactions on Electrical Energy Systems Start Football data stream Improved Openpose network framework Test Excellent Improve the model No network performance evaluation Yes Human body key point coordinate data Neural network framework Test Excellent model Improve the No performance network evaluation Yes Limb movement analysis Algorithm transformation method Finish Figure 5: Flow chart of football action analysis model design. 2.3. Football Player Training and Teaching Actions Based on (2) Deep Belief Network Artiﬁcial Intelligence DBN algorithm is a kind of neural network of (1) With the help of artiﬁcial intelligence for football machine learning. It can be used for unsupervised players’ training and teaching actions, detailed plan learning and supervised learning. DBN is a design and implementation can be carried out. probabilistic generative model. Compared with the Under this technology, it can be simply divided into traditional neural network of discriminant model, the following modules [21]. First, the relevant data of the generative model is to establish a relationship football players are entered into the server, and the between the observation data and the label. For the topological structure is managed according to the input feature data, usually the data are ﬁtted with a relevant operation of the server, as shown in Figure 6. presumption of what distribution it conforms to, and then, it is trained and solved according to its As shown in Figure 6, the server is used as a node to assumed distribution model. Many distributions manage diﬀerent related data of athletes, and the can take advantage of the unique properties and logical processing server of the system structure is learning process of the energy model, limiting the expanded into athlete management, athlete-related Boltzmann machine [22, 23] to be transformed data, trainer management, and background man- from the energy model. *en, the energy deﬁnition agement and is timely feedback based on diﬀerent is as follows: changes each time to improve the athlete’s training. International Transactions on Electrical Energy Systems 7 obtain from the image. *e global information of the data in Sports related the image is extracted, and a global feature extraction model data Manager [25, 26]isestablished. In action recognition, the position management of changes in key parts of the human body in the action se- course training quence are generally used for estimation. *e speciﬁc Training plan management training management system for the human body is shown in Figure 7. 3. Experiment and Analysis Backstage Athlete management Management 3.1. Data Analysis of Artiﬁcial Intelligence. Although Serverice movement is an external manifestation of the ability to collaborate, the completion of technical movement is the Figure 6: Schematic diagram of the topological structure of result of the coordinated work of nerves, muscles, and football player training management. sensations. According to the related theory of cooperation ability, the football player’s physical cooperation ability is deﬁned. In the understanding of artiﬁcial intelligence, it n m n m mainly relies on receipt data for timely feedback. For the E(k|θ) � 􏽘 a v − 􏽘 b h − 􏽘 􏽘 v h . (14) i i j j i j understanding of our neural network, we must also rely on i�1 j�1 i�1 j�1 the calculation of the network and the establishment of the model. In-depth understanding of various data can be ad- Among them, θ � [a , b ] is the parameter model of the i j justed in various physical indicators [27]. Table 1 shows the following formula: impact of diﬀerent groups of people on the impact of the − E(h|θ) introduction of artiﬁcial intelligence technology in football. (15) K(h|θ) � . From the data in Table 1, we can see whether the impact P(θ) on football players after the introduction of artiﬁcial in- Among them, P(θ) refers to the normalization factor in telligence will increase the interest of the players, and most the calculation of joint probability: importantly, it will bring about huge changes in improving − E(h|θ) the training quality of the players. *rough the investi- P(θ) � 􏽘 e . (16) gation of a large number of samples, it is found that there h�1 will be an 85% impact on improving the quality of training. *e establishment of the energy function [24] is to obtain Of course, there are other impacts similar to the use of the distribution of the input data, the input data can see the artiﬁcial intelligence technology to record relevant data of state of the layer unit, and the likelihood function is solved athletes and then provide better feedback based on the through speciﬁc calculations. *e formula can be expressed content obtained. After receiving the relevant feedback, the as follows: athlete can adjust the relevant training plan in time. *en, after receiving the relevant feedback, the athlete can adjust − E(v,h|θ) 􏽐 e (17) the relevant training plan in time. In addition to the rel- P(v, h|θ) � . Z(θ) evant changes in the training, the athlete can adjust the action in time during the training to reduce the physical Knowing the input data of the RBM, you can see the state injury. of the layer unit. Because of the structural characteristics of no connection in the layer and full connection between the layers, the formula for obtaining the state of the hidden layer 3.2. Training Behavior of Football Players. After the intro- unit is as follows: duction of artiﬁcial intelligence, athletes will reduce their usual mistakes, etc. *e impact of improving on the basis of ⎣ ⎦ ⎡ ⎤ P v � 1|v, θ􏼁 � μ a + 􏽘 v . (18) i i i existing levels by trainers is shown in Table 2. According to the data in the table, the inﬂuence of the research objects of diﬀerent competition levels can be ob- In the formula, μ is the activation function. According to served. First of all, certain research is conducted on whether the state of the hidden layer unit, the formula for obtaining the changes brought about by the introduction of artiﬁcial the visible layer unit in the reverse direction is as follows: intelligence will have an impact. Experimental data show that after the introduction of artiﬁcial intelligence, the ⎢ ⎥ ⎡ ⎢ ⎤ ⎥ ⎣ ⎦ P v � 1|h, θ􏼁 � μ b + 􏽘 h . (19) i j j changes brought about by both professional leagues and campus leagues are very large. For example, for school In motion recognition, motion data are a kind of real- leagues, the main impact is about 30%. In addition, the time data. *e image-based representation method means changes brought by the training team members introducing that the original data come from the image, and the relevant artiﬁcial intelligence technology for training are shown in model is established to extract the information we want to Table 3. 8 International Transactions on Electrical Energy Systems Training record Physical Trainer Training trainer Physical Adjust the Training Formulate plan Record Manage Player Belong Figure 7: Football player training management system database. Table 1: Surveyed football players on the impact of the introduction of artiﬁcial intelligence technology. Percentage Projects Number of samples (%) Increase interest in training 1008 83 Improve training quality 2013 85 Reduce injuries 1111 82 More comfortable to wear 1018 25 Complete training plan better 1339 52 Table 2: Level of the football players surveyed and the impact of the introduction of artiﬁcial intelligence. Level of inﬂuence Business league School league College league Professional league Percentage Very small 1600 1220 3311 0 10% Smaller 0 1302 4025 3000 25% Bigger 2221 1400 2030 2845 75% Very big 2012 1502 2050 3102 0 Percentage 42% 30% 10% 18% 100% Table 3: About the survey results after the use of artiﬁcial intelligence equipment in auxiliary training competitions in my country. Select item Number of surveys Survey percentage (%) Better reﬂect the problems that arise in the game 500 100 Improve players’ motivation for training 300 45 Make better training plans 130 72 Improve training eﬃciency 300 20 Reduce athletes’ injuries 90 15 Table 4: Athlete detection statistics after the adversarial personal improve the training eﬃciency of football players and ball control test experiment. improve the enthusiasm of the players in training. In addition to the intensity of the original training, it can MSQ Test group Control group Sample T value also reduce the chance of athletes being injured. K1 8.26± 1.13 8.45± 3.21 0.982 0.028 *rough the adjustment of big data, certain feedback K2 9.32± 2.21 8.23± 0.88 1.103 0.037 from athletes can be obtained. In turn, the formation of a K3 7.28± 2.28 6.63± 1.48 0.656 0.037 good training closed-loop promotion can not only improve the training eﬃciency of the athletes but also According to Table 3, it can be seen that artiﬁcial help the athletes to increase their interest in many as- intelligence equipment has a great auxiliary eﬀect on pects [28]. football. After using artiﬁcial intelligence, based on the *rough the above table method, the introduction of related training of the original football players, it can artiﬁcial intelligence in football can bring about very big International Transactions on Electrical Energy Systems 9 11 11 11 11 11 11 11 10 10 10 10 10 10 10 10 10 10 9 9 99 9 9 9 9 9 9 8 8 8 8 88 8 10 8 8 8 8 8 8 7 7 7 7 7 66 66 666 6 123456789 10 11 12 123456789 10 11 12 Test group (a) Test group (c) Test group (b) Test group (d) Control group Control group Figure 8: Comparison of capacity changes before and after getting rid of the connection test. Sliding size 5 101520253035 Accuracy Score (1) Score (2) Figure 9: Selection of diﬀerent frame numbers for sliding window. changes. It is not only reﬂected in the behavioral teaching of experimental group, the lowest number of successes is 7 football players’ training but also can better stimulate the times, and the lowest number of successes in the control potential of the players when playing games and improve the group is 6 times. *e comparison shows that the number of physical performance to a new height. At the same time, it is successes of the experimental group is signiﬁcantly better than that of the control group. *erefore, according to the not only reﬂected in sports competitions but also in pro- tecting the athletes’ bodies to ensure that they will not be results of this experiment, it can be seen that the intro- harmed. *is timely feedback can help trainers and athletes duction of artiﬁcial intelligence technology can improve the to collaborate better. athlete’s antagonistic personal ball control ability. From the perspective of diﬀerent experimental groups and control groups, there is a gap in the number of athletes’ 3.3. Experimental Football Players’ Ability Test Results after confrontational escape transfers. Observing the data in the Artiﬁcial Intelligence Training. After the experiment, the chart, we can see that the data of the experimental group are analysis of the test results of football players’ personal ball more stable, and at the same time, it is better than the control is to detect the antagonistic personal ball control of experimental group, as shown in Figure 8. diﬀerent groups of players using a paired sample test. K1 is Because football is a long-term sequence analysis [29] the preexperimental test result of the number of successful process, it is necessary to use a sliding window to segment transfers after the athlete gets rid of the test, K2 is the test diﬀerent data. *erefore, to ﬁnd the appropriate length of result of the number of athletes successfully shed the ball in the sliding window, the accumulated coordinates in diﬀerent the test before the test, and K3 is the test result of the athlete’s frames are obtained, as shown in Figure 9. catch-and-stop quality before the test in the test, as shown in Table 4. 4. Discussion From Table 4, it is observed that in the test about athletes’ antagonistic personal ball control, the number of successes *is article is dedicated to introducing artiﬁcial intelligence after the athletes get rid of the number of transfer experi- technology into the teaching behavior of football sports ments is basically between 10. According to the data in the training. *e research in the article shows that when 10 International Transactions on Electrical Energy Systems [6] M. Seyedmahmoudian, B. Horan, T. K. Soon et al., “State of computer technology and football are combined, certain the art artiﬁcial intelligence-based MPPT techniques for advantages can be obtained in the individual physical in- mitigating partial shading eﬀects on PV systems—a review,” dicators of athletes, and better feedback can be obtained Renewable and Sustainable Energy Reviews, vol. 64, pp. 435– during training to form a closed-loop virtuous circle. 455, 2016. *rough the analysis of diﬀerent frame numbers, the ath- [7] M. Wang and Z. Liang, “An extraction method of volleyball lete’s range of activities is captured, and then, quantitative spiking trajectory and teaching based on wireless sensor analysis is carried out, so that football training can get rid of network,” Security and Communication Networks, vol. 2021, purely relying on experience and enter the era of theory and no. 9, Article ID 9966994, 11 pages, 2021. digitalization. [8] V. Patel and V. Savsani, “Multi-objective optimization of a Stirling heat engine using TS-TLBO (tutorial training and self learning inspired teaching-learning based optimization) al- 5. Conclusions gorithm,” Energy, vol. 95, pp. 528–541, 2016. *e data acquired by each device can simulate the entire [9] H. Lu, Y. Li, M. Chen, H. Kim, and S. Serikawa, “Brain in- process of the athlete from before touching the ball to after telligence: go beyond artiﬁcial intelligence,” Mobile Networks and Applications, vol. 23, no. 2, pp. 368–375, 2017. ejection. *e height and distance of the virtual football ﬂight [10] X. Liang, L. Haiping, J. Liu, and L. Lin, “Reform of English are simulated according to the football players’ exertion habits interactive teaching mode based on cloud computing artiﬁcial and contact points during sports. Artiﬁcial intelligence intelligence—a practice analysis,” Journal of Intelligent and technology actually simulates the physical characteristics of Fuzzy Systems, vol. 40, no. 2, pp. 3617–3629, 2021. football players’ height, body shape, etc., according to their [11] B. Li, Y. Fei, and H. Liu, “An artiﬁcial intelligence based model joints and their spatial positions, and appears in the simulated for evaluation of college students’ ability and characteristics environment at a speciﬁc speed. *rough the introduction of through teaching evaluation,” Journal of Intelligent and Fuzzy artiﬁcial intelligence technology, the use of multi-person Systems, vol. 40, no. 2, pp. 3397–3407, 2021. whole-body positioning and interaction systems, and online [12] Z. Chen, Y. Lian, and Z. Lin, “Research on business English training, you can see the actions of other operators and load autonomous learning based on artiﬁcial intelligence and and update the training scene in real time. Football training or improved BP network model,” Journal of Intelligent and Fuzzy games can be carried out online, which can help users visually Systems, vol. 40, no. 3, pp. 1–12, 2020. analyze the on-site situation and make correct decisions about [13] L. Yu and N. Peng, “Research on English teaching reform based on artiﬁcial intelligence matching model,” Journal of passing height, angle, and time. Intelligent and Fuzzy Systems, no. 1, pp. 1–10, 2021. [14] G. Yunfeng and L. Jing, “Recognition of teaching method Data Availability eﬀects based on grid model simpliﬁcation and artiﬁcial in- telligence,” Journal of Intelligent and Fuzzy Systems, vol. 40, *e data that support the ﬁndings of this study are available no. 4, pp. 6707–6717, 2021. from the corresponding author upon reasonable request. [15] Z. Zhang, “Research and application of capability evaluation model based on BP neural network educational tech- nology—taking the problem-based learning teaching process Conflicts of Interest learning as an example,” Journal of Computational and *e authors declare that there are no conﬂicts of interest <eoretical Nanoscience, vol. 13, no. 9, pp. 6210–6217, 2016. with respect to the research, authorship, and/or publication [16] S. Zou, “Designing and practice of a college English teaching platform based on artiﬁcial intelligence,” Journal of Compu- of this article. tational and <eoretical Nanoscience, vol. 14, no. 1, pp. 104– 108, 2017. References [17] Y. Ding, “Performance analysis of public management teaching practice training based on artiﬁcial intelligence [1] D. Hassabis, D. Kumaran, C. Summerﬁeld, and M. Botvinick, technology,” Journal of Intelligent and Fuzzy Systems, vol. 40, “Neuroscience-inspired artiﬁcial intelligence,” Neuron, no. 2, pp. 3787–3800, 2021. vol. 95, no. 2, pp. 245–258, 2017. [18] J. Zhang and K. B. Letaief, “Mobile edge intelligence and [2] L. D. Raedt, K. Kersting, S. Natarajan, and D. Poole, “Sta- computing for the internet of vehicles,” Proceedings of the tistical relational artiﬁcial intelligence: logic, probability, and IEEE, vol. 99, pp. 1–16, 2019. computation,” Synthesis Lectures on Artiﬁcial Intelligence and [19] N. Dhieb, H. Ghazzai, H. Besbes, and Y. Massoud, “A secure Machine Learning, vol. 10, no. 2, pp. 1–189, 2016. AI-driven architecture for automated insurance systems: [3] L. Rongpeng, Z. Zhao, X. Zhou et al., “Intelligent 5G: when fraud detection and risk measurement,” IEEE Access, vol. 8, cellular networks meet artiﬁcial intelligence,” IEEE Wireless pp. 58546–58558, 2020. Communications, vol. 24, no. 5, pp. 175–183, 2017. [20] X. Qiao, Y. Huang, S. Dustdar, and J. Chen, “6G vision: an AI- [4] J. H. *rall, X. Li, Q. Li et al., “Artiﬁcial intelligence and driven decentralized network and service architecture,” IEEE machine learning in radiology: opportunities, challenges, Internet Computing, vol. 24, no. 4, pp. 33–40, 2020. pitfalls, and criteria for success,” Journal of the American [21] D. Y. Chao, “Computation of elementary siphons in petri nets College of Radiology, vol. 15, no. 3, pp. 504–508, 2018. for deadlock control,” <e Computer Journal, vol. 49, no. 4, [5] P. Glauner, J. A. Meira, P. Valtchev, R. State, and F. Bettinger, “*e challenge of non-technical loss detection using artiﬁcial pp. 470–479, 2006. [22] S. M. Senouci, H. Sedjelmaci, J. Liu, M. H. Rehmani, and intelligence: a survey,” International Journal of Computational Intelligence Systems, vol. 10, no. 1, pp. 760–775, 2017. E. Bou-Harb, “AI-driven cybersecurity threats to future International Transactions on Electrical Energy Systems 11 networks [from the guest editors],” IEEE Vehicular Tech- nology Magazine, vol. 15, no. 3, pp. 5-6, 2020. [23] K. Karras, E. Pallis, G. Mastorakis et al., “A hardware ac- celeration platform for AI-based inference at the edge,” Circuits, Systems, and Signal Processing, vol. 39, no. 2, pp. 1059–1070, 2020. [24] Y. Zhou and X. Zhu, “Analysis of vehicle network architecture and performance optimization based on soft deﬁnition of integration of cloud and fog,” IEEE Access, vol. 7, pp. 101171–101177, 2019. [25] D. Zeng, L. Gu, S. Guo, Z. Cheng, and S. Yu, “Joint opti- mization of task scheduling and image placement in fog computing supported software-deﬁned embedded system,” IEEE Transactions on Computers, vol. 65, no. 12, pp. 3702– 3712, 2016. [26] Y. Ramzanpoor, M. Hosseini Shirvani, and M. Golsorkhtabaramiri, “Multi-objective fault-tolerant opti- mization algorithm for deployment of IoT applications on fog computing infrastructure,” Complex & Intelligent Systems, vol. 8, no. 1, pp. 361–392, 2021. [27] M. K. Hussein and M. H. Mousa, “Eﬃcient task oﬄoading for IoT-based applications in fog computing using ant colony optimization,” IEEE Access, vol. 8, pp. 37191–37201, 2020. [28] C. Guerrero, I. Lera, and C. Juiz, “A lightweight decentralized service placement policy for performance optimization in fog computing,” Journal of Ambient Intelligence and Humanized Computing, vol. 10, no. 6, pp. 2435–2452, 2019. [29] T. Huang, W. Lin, C. Xiong, R. Pan, and J. Huang, “An ant colony optimization-based multiobjective service replicas placement strategy for fog computing,” IEEE Transactions on Cybernetics, vol. 51, pp. 1–14, 2020.

Journal

International Transactions on Electrical Energy Systems – Hindawi Publishing Corporation

Published: Aug 30, 2022

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Evaluation Model of Football Players’ Training and Teaching Actions Based on Artificial Intelligence

Evaluation Model of Football Players’ Training and Teaching Actions Based on Artificial Intelligence

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Evaluation Model of Football Players’ Training and Teaching Actions Based on Artificial Intelligence

Evaluation Model of Football Players’ Training and Teaching Actions Based on Artificial Intelligence

References (30)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies