## Assessment of high-gain quadratic boost converter with hybrid-based maximum power point tracking technique for solar photovoltaic systems

**Abstract**

Abstract Solar photovoltaic (SPV) modules have a low output voltage and are load-dependent. Therefore, it is critical that the SPV system has an adequate DC–DC converter to regulate and improve the output voltage to get maximum output voltage. To meet load requirements, the voltage must be increased, necessitating the use of energy-efficient power electronic converters. The performance of an SPV system coupled to a high-gain quadratic boost converter (HG-QBC) with a load is investigated in this paper. The suggested HG-QBC for the SPV system at a lower value of duty ratio provides high voltage gain with a boost factor of four times. An analytical comparison is carried out with the various existing boost converters in terms of the components and the boost factor. The issue of locating the maximum power generation point from the SPV system is crucial. As a result, choosing an appropriate maximum power point tracker (MPPT)-based technique to obtain the peak power output of the SPV system under the rapidly varying atmospheric conditions is vital. To determine the highest output power of an SPV system, a hybrid-based MPPT with a neural network assisted by a perturb and observe (P&O) technique is proposed. For the HG-QBC, a comparison of the proposed MPPT with a traditional P&O-based MPPT is illustrated. The comparative analysis takes into account rise time, settling time and voltage ripples. The output voltage and power characteristics of the proposed model are analysed under constant and varying irradiation conditions using MATLAB®/Simulink®. The results of a hybrid-based MPPT show that the oscillations are minimum at the maximum power point with fewer ripples of 0.20% and a settling time of 1.2 s in comparison with the other two techniques. Open in new tabDownload slide neural network, maximum power point tracker, renewable energy sources, power electronic converters, solar photovoltaic and high-gain quadratic boost converter Introduction Due to emissions of greenhouse gases and the energy crisis, the use of fossil fuels in electrical power generation is being phased out in favour of renewable energy sources (RES). RES currently account for 25% of worldwide power generation capacity. The world’s population is expanding, the environment is deteriorating due to the consumption of fossil fuels and electricity must be transported to distant locations; these are all reasons driving the usage of RES today. The rise in facilities for renewable energy is primarily due to excess pollution levels produced by the use of fossil fuels to meet rising electrical demand [1–3]. The maintenance required for RES is low, they are environmentally friendly and they are very efficient as well. The power generation from a solar photovoltaic (SPV)-based RES system can alleviate energy crises and reduce environmental pollution. SPV power has been increasing rapidly and is predicted to increase 3-fold in the near future. The limitations of the SPV system are that the output voltage is load-dependent, small in magnitude and affected by environmental conditions including solar irradiation and temperature. SPV systems are set in a series–parallel combination to reach high output voltage levels, which results in a reduction in efficiency and a large system size [4]. As a result, any SPV system must incorporate an adequate DC–DC converter to boost and vary the output voltage (Vout) because voltage must be increased to meet load requirements. Therefore, an energy-efficient power electronic converter is required. The different types of DC–DC converters are investigated in [5–9]. All these converters have a few limitations, such as: high stress on switching devices; conversion efficiency is low; the system is bulky; and the output voltage cannot be enhanced multiple times. As a result, a high-gain boost converter (BC) can maximize the signal to considerably higher levels than at the input by altering the duty cycle. High voltage gain is provided by quadratic boost converters (QBCs), especially at low duty ratios for SPV applications. Apart from the high voltage gain, the QBC topology decreases voltage stress on the switch, diodes and capacitors, which is another benefit [10]. It is critical to incorporate a suitable DC–DC BC with high voltage gain to operate the SPV system at the maximum power point (MPP) under rapidly changing atmospheric weather conditions. An efficient power converter with a suitable MPP tracking (MPPT) system gives a boost to the efficiency with reduced conversion losses [11]. An MPPT for a DC–DC converter can be implemented using a control algorithm and embedded system. It is built into the SPV system to maximize energy harvesting in real time under a variety of operating and environmental conditions. The use of an artificial neural network (NN)-based MPPT has the advantage of allowing the MPP to be tracked quickly and accurately. The effectiveness of the hidden-layer algorithm and the NN’s training is determined by these factors [12]. The MPPT alters the converter to match the impedance of the SPV system with the load. Fig. 1 shows an SPV system coupled to a load via an MPPT. Fig. 1: Open in new tabDownload slide SPV system coupled to a load via MPPT. MPPT controllers share several fundamental characteristics, including their simplicity of design, low cost, high-performance characteristics with minimal output power fluctuation and the ability to track efficiently and rapidly under changing operating conditions. Tracking the point of maximum power generation from the SPV system is a critical task. Hence, it is important to select a suitable MPPT-based technique to harvest maximum power from the SPV under sudden change in weather conditions. The different MPPT techniques have been investigated in [13–18]. The conventional techniques include constant voltage, perturb and observe (P&O), hill climbing, incremental conductance (INC), etc., and soft computing-based techniques include fuzzy logic controller, bat algorithm (BA), particle swarm optimization, NN, etc. There are several ways for controlling the duty cycle of DC–DC converters [19–22] to maximize the output power (Pout) of SPV systems under all weather conditions. Among the numerous MPPT methods available, the P&O technique is used due to its popularity, accuracy, simplicity of design and ease of application [23–28]. The P&O approach focuses on perturbing the SPV system voltage to optimize the DC–DC converter’s duty cycle and operate under MPP. But it does not perform well in rapidly changing environmental conditions and due to this, the SPV system efficiency is reduced. As a result, before employing the P&O approach, it is required to concentrate on improving the performance of the conventional approach. Therefore, soft computing-based approaches are adapted to track the global MPP more efficiently. In this paper, a hybrid-based MPPT with NN assisted by the P&O technique is proposed due to its dynamic behaviour under sudden changing atmospheric conditions. The tracking control and stability of the NN-based MPPT technique are inherently robust against internal system parameters and load uncertainties. This paper presents an assessment of a solar-powered QBC to achieve high voltage gain and compares it with the traditional BC. The performance of the developed hybrid-based MPPT with NN assisted by the P&O technique has been analysed using the MATLAB®/Simulink® tool for constant and varying irradiation levels. 1 SPV system The electrical equivalent circuit of an SPV system is depicted in Fig. 2. The equivalent circuit includes a photocurrent source (Iph), a series resistor (Rs), a parallel resistor (Rsh) and a diode (D). The mathematical modelling of the SPV system is discussed in detail [29]. The expression for the current of a SPV cell is obtained by applying Kirchhoff’s current law: Fig. 2: Open in new tabDownload slide Electrical equivalent circuit of an SPV cell. Iph= IPV+ID+Ish(1) IPV=Iph−ID−Ish(2) where IPV represents the output of the SPV cell, Iph represents the photon current, ID represents the diode current and Ish represents the current flowing through Rsh. 1.2 Conventional boost converter The conventional boost converter (CBC) is depicted in Fig. 3. The output of the CBC is very low to supply the load and the boost factor (B) is limited to 2. The low value of the boost factor is the main demerit of the CBC. The voltage gain or boost factor of the CBC is given by Fig. 3: Open in new tabDownload slide Conventional boost converter. Vout VPV = 1(1−D)(3) Equation (3) represents the voltage gain of the CBC. The value of B for a duty cycle of 0.5 is 2. The limitations of the CBC can be overcome with the use of a high-gain quadratic boost converter (HG-QBC). 2 Operation of an HG-QBC The proposed HG-QBC is depicted in Fig. 4 and consists of a semiconductor switch (S), three diodes (D1, D2 and D3), two capacitors (C1 and C2), two inductors (L1 and L2) and finally the load resistance (R). The proposed converter uses a single switch (S) to control the output voltage. Fig. 4: Open in new tabDownload slide Proposed HG-QBC connected to SPV. 2.1 Mode-I operation (0 to DT) The proposed QBC for mode-I operation is depicted in Fig. 5. In this mode, the switch (S) will be turned on. The diode (D1) will be in conduction and the other two diodes (D2 and D3) will become open-circuit. The inductor (L1) is charged by a photovoltaic source (VPV), the inductor (L2) is energized by the total voltage supplied by the capacitor (C1) and the capacitor (C2) provides current to the load resistor (R). The voltage across the inductors (L1 and L2) is obtained by applying Kirchhoff’s voltage law (KVL) to Fig. 5: Fig. 5: Open in new tabDownload slide Mode-I operation of proposed converter. VL1=VPV(4) VL2=VC1(5) 2.2 Mode-II operation (DT to T) The mode-II operation of the suggested QBC is depicted in Fig. 6. The switch (S) will be deactivated. The diode (D1) will be reverse-biased, while the other two diodes (D2 and D3) will be forward-biased and in conduction. In this mode, the L1 and L2 will be demagnetized, and C1 and C2 will be charged. Fig. 6: Open in new tabDownload slide Mode-II operation of proposed converter. The voltage across the inductors (L1 and L2) is obtained by applying KVL to Fig.6: VL1=VPV−VC1 (6) VL2=VC1−VC2(7) VL2=VC1−Vout(8) 2.3 Analysis of the HG-QBC The voltage across L1 is obtained by applying a volt–sec balance: ∫T0VL1(t)dt=0(9) ∫DT0VPVdt+∫TDT(VPV−VC1)dt=0(10) VPV[t]0DT+(VPV−VC1)[t]0DT(11) The voltage across the capacitor C1 is given as: VPV=VC1(1−D)(12) VC1=VPV(1−D)(13) Volt–sec balance is used to determine the voltage across L2: ∫T0VL2(t)dt=0(14) ∫DT0VC1dt+∫TDT(VC1−Vout)dt=0(15) VC1[t]0DT+(VC1−Vout)[t]0DT(16) The voltage across the capacitor C2 is given as: VC1=Vout(1−D)(17) Vout=VC1(1−D)= VPV(1−D)2(18) The voltage gain of the HG-QBC is given by: VoutVPV= 1(1−D)2(19) Equation (19) represents the voltage gain of the HG-QBC. For a 0.5 duty cycle (D), B is 4. In comparison to the CBC for the SPV system, the B-value is high. Table 1 shows a detailed comparison of different existing BCs [13–17] based on the number of components and the boost factor B. It reveals that all of the converters mentioned have a poor value of B when compared to the HG-QBC. To combat the demerits of the CBC, an HG-QBC is used to integrate the SPV system into the load. The Vout of the SPV system is boosted to four times the input voltage and fed to the load. The proposed HG-QBC and converter E have the same value of B but in terms of component count, the proposed converter has only 8 components and converter E has a total of 12 components, which is an advantage of the HG-QBC. Table 1: A comparative analysis of various boost converters at 0.5 duty cycle Type of BC . Gain . Components . . . . Total . B . . . S . D . L . C . . . CBC 11−D 1 1 1 1 4 2 Converter A [13] 1+D1−D 1 1 2 3 7 3 Converter B [14] 1+D1−D 1 3 2 3 9 3 Converter C [15] 2 D1−D 1 2 2 3 8 2 Converter D [16] 2−D1−D 1 2 1 2 6 3 Converter E [17] 1+3D1−D 1 3 4 4 12 4 HG-QBC (proposed) 1(1−D)2 1 3 2 2 8 4 Type of BC . Gain . Components . . . . Total . B . . . S . D . L . C . . . CBC 11−D 1 1 1 1 4 2 Converter A [13] 1+D1−D 1 1 2 3 7 3 Converter B [14] 1+D1−D 1 3 2 3 9 3 Converter C [15] 2 D1−D 1 2 2 3 8 2 Converter D [16] 2−D1−D 1 2 1 2 6 3 Converter E [17] 1+3D1−D 1 3 4 4 12 4 HG-QBC (proposed) 1(1−D)2 1 3 2 2 8 4 S, switches; D, diodes; L, inductors; C, capacitance. Open in new tab Table 1: A comparative analysis of various boost converters at 0.5 duty cycle Type of BC . Gain . Components . . . . Total . B . . . S . D . L . C . . . CBC 11−D 1 1 1 1 4 2 Converter A [13] 1+D1−D 1 1 2 3 7 3 Converter B [14] 1+D1−D 1 3 2 3 9 3 Converter C [15] 2 D1−D 1 2 2 3 8 2 Converter D [16] 2−D1−D 1 2 1 2 6 3 Converter E [17] 1+3D1−D 1 3 4 4 12 4 HG-QBC (proposed) 1(1−D)2 1 3 2 2 8 4 Type of BC . Gain . Components . . . . Total . B . . . S . D . L . C . . . CBC 11−D 1 1 1 1 4 2 Converter A [13] 1+D1−D 1 1 2 3 7 3 Converter B [14] 1+D1−D 1 3 2 3 9 3 Converter C [15] 2 D1−D 1 2 2 3 8 2 Converter D [16] 2−D1−D 1 2 1 2 6 3 Converter E [17] 1+3D1−D 1 3 4 4 12 4 HG-QBC (proposed) 1(1−D)2 1 3 2 2 8 4 S, switches; D, diodes; L, inductors; C, capacitance. Open in new tab The boost factors for various converters are plotted in Fig. 7. The value of B is higher for the proposed HG-QBC than for other converters. Fig. 7: Open in new tabDownload slide Boost factors of various BCs. 3 MPPT A vital component of an SPV system is the MPPT, which allows maximum power to be harvested at the MPP [30–32]. 3.1 P&O MPPT For computing the MPP, this technique is quite simple and easy to implement, as it only requires measuring the current and voltage from the SPV module. It is based on an iterative procedure for tracing out the SPV’s MPP. In this method, a slight perturbation is introduced to cause the PV module’s power to alter. The output power of the SPV module is monitored regularly and analysed with the past value. By varying the duty cycle of the BC, the Vout of the SPV is either enhanced or reduced to compute the MPP. After perturbation, the voltage increases leading to more power than the past value. The process recaps and adjusts the duty cycle in an ascending direction to determine an enhanced MPP. The duty cycle is decreased to reduce the perturbation, once the power generated is less than the past value, and the procedure is continued until MPP is obtained. The voltage is always fluctuating in a steady state that causes ripples in the Vout and finally the power loss. The P&O MPPT has a major flaw in this regard. Larger perturbations produce a faster dynamic response, but they also produce more oscillations at the output. Smaller perturbations have a slower dynamic response but fewer oscillations at steady state. The P&O MPPT algorithm flow chart is depicted in Fig. 8. Fig. 8: Open in new tabDownload slide Flowchart of P&O MPPT algorithm. Now, in terms of the algorithm, first the PV panel’s operational voltage is changed by a small amount and the resulting power (P) is measured. If P changes in a positive direction, the current operating point is getting closer to the MPP and hence more variation is made in that direction. The operational point is presumed to be a departure from the MPP. Power has been shifted in a negative direction and the direction of perturbation must be reversed to approach the MPP. This method does not take account of rapid changes in the irradiation level and due to perturbation, it changes the MPP and ends up with a wrong MPP. The benefits of using the P&O method are that it has a simple algorithm, is easy to implement, has a lower implementation cost compared to other methods and is more accurate than other methods. But this method has a few limitations, such as, under the steady-state condition, the output power oscillates at MPP and the operation speed is slower. In order to overcome these limitations, the NN-based MPPT method is used. The proposed hybrid MPPT is analysed with the conventional P&O and NN MPPT at constant and various solar irradiation conditions. 4 NN MPPT An NN is first trained using a similar data set and is then used or tested on a different set of data. A multilayer feed-forward NN is chosen to implement MPPT techniques for SPV applications. NNs are mainly categorized into three layers, such as input layers, hidden layers and output layers. The multilayer feed-forward NN has two different phases: (i) Training phase (Learning phase): it works by providing a specific input to the NN to get a specific output. This can be done through continuous training on a set of training data. (ii) Execution phase: the outputs are returned based on the value of the input. The demerits of P&O-based MPPT can be overcome using an NN-based MPPT controller due to its advantages such as adaptive learning, fault tolerance capability, real-time operation and self-organization where the NN forms its own structure that is suitable for processing the information received from the training data set. The input layer will receive the inputs from the SPV model such as cell temperature (T), solar irradiance (G) and photovoltaic voltage (VPV). Then it passes through all the hidden layers and finally it returns to the output layer as photovoltaic current (IPV). In order to obtain the output, it is easy to propagate an input through the network of a feed-forward NN. The feed-forward NN-based MPPT architecture for the SPV model is depicted in Fig. 9. It considers three inputs (T, G and VPV) and one output (IPV). Fig. 9: Open in new tabDownload slide NN-based MPPT architecture for SPV model. The NN model’s output current is given by: IPV=net(u)=W2tanh(W1u+b1)+b2(20) where u=[T G V]T W1, W2, b1 and b2 are the set of coefficients that must be realized from the input/output training sets that are available. Each input can be standardized using the formula to help with the learning process: ui,new=ui−ui,minui,max−ui,min, i∈{ T, G, V}(21) where index max/min denotes the maximal/minimal values of T, G and V in the training set. The Levenberg–Marquardt algorithm is used to optimize the NN coefficients. The cost function is the mean square difference between the expected current and the actual current. The equation for the cost function is: J=1∑i=1MNi∑i=1M∑k=1Ni(IPV−IPV)2(22) Fig. 10 shows that the feed-forward NN is used to predict the MPP voltage (VMPP). The cell temperature (T) and solar irradiance (G) are used to measure and fed to the input layer. The NN-based MPP voltage estimator will provide VMPP instantaneously under fast-changing atmospheric weather conditions. The irradiance information will not be predicted, which is a disadvantage of this algorithm. Fig. 10: Open in new tabDownload slide NN-based MPP voltage estimator. Fig. 11 shows that the feed-forward NN is used to estimate the irradiance. The SPV current and voltage are measured and fed to the input layer. The advantage of this is that only a few additional computational operations are required. The proposed estimator is based on the principle of immersion and invariance [16]. If IPV and T are both measurable, then the difference between the measured and predicted current of the SPV module is given by: Fig. 11: Open in new tabDownload slide NN-based irradiance estimator. ek= IPV−IPV=Ik−net(Gk,Tk Vk)(23) The immersion and invariance principle of the estimator is given as: G(k+1)=G(k)+γ e(k)(24) With increasing irradiance, the current in the SPV module increases. 4.1 NN-based MPPT algorithm The iterative, analytical rules for calculating MPP voltage and estimating irradiance are derived from the NN model’s output current expression. Although the NN-based MPPT controller cannot estimate the ideal voltage in real time, it has adjustable parameters that may be used to control the SPV, tracking speed and calculation complexity. The relationship between voltage, current and output power is represented in Equation (25): P=V×I=V×I(G,T,V)(25) The output power at MPP is given in Equation (26): PMPP=P(VMPP,IMPP)≥P(V,I) ∀ V,I(26) The MPP of the SPV module can be determined using the gradient ascent rule: Vk+1=Vk+μ∂Pk∂Vk=Vk+μ∂(Vk×Ik)∂Vk(27) Vk+1=Vk+μ (Ik+Vk×∂Ik∂Vk)(28) where k represents the instantaneous discrete time and μ represents the convergence speed controlled by a positive step size. For larger step sizes, the NN-based MPPT algorithm will converge faster. The SPV Vout and current are fed to the input layer. The voltage and current from the SPV panel are sensed by the NN-based MPPT controller. Then it detects the change in irradiance (G). After detecting any abrupt irradiance change, the NN-based MPPT controller will measure the voltage and power at MPP. The NN-based MPP voltage estimator will provide VMPP instantaneously under the fast-changing atmospheric weather conditions. The irradiance information will not be predicted, which is a disadvantage of this algorithm. The flowchart of NN- and hybrid-based MPPT techniques for SPV systems are depicted in Figs 12 and 13, respectively. Fig. 12: Open in new tabDownload slide Flowchart of NN-based MPPT controller. Fig. 13: Open in new tabDownload slide Flowchart of hybrid-based MPPT for SPV. 5 Results and discussion The simulation studies are carried out with the help of the MATLAB®/Simulink® software. In this paper, the Kyocera Solar KC200GT module is considered with a constant irradiation (G) of 1000 W/m2. The MATLAB®/Simulink® tool is used to simulate the chosen module for various values of G to obtain the output characteristics. The various parameters used for simulation studies are tabulated in Table 2. Table 2: Solar-powered HG-QBC parameters Parameters . Values . Pmax (maximum power) 200 W Voltage at MPP 26.2 V VPV to HG-QBC 26.2 V Isc (short-circuit current) 8.21 A Voc (open-circuit voltage) 32.9 V Duty cycle 0.5 Switching frequency 50 kHz Inductors L1 = 0.2 mH L2 = 0.5 mH Capacitors C1 = 9000 µF C2 = 2000 µF Parameters . Values . Pmax (maximum power) 200 W Voltage at MPP 26.2 V VPV to HG-QBC 26.2 V Isc (short-circuit current) 8.21 A Voc (open-circuit voltage) 32.9 V Duty cycle 0.5 Switching frequency 50 kHz Inductors L1 = 0.2 mH L2 = 0.5 mH Capacitors C1 = 9000 µF C2 = 2000 µF Open in new tab Table 2: Solar-powered HG-QBC parameters Parameters . Values . Pmax (maximum power) 200 W Voltage at MPP 26.2 V VPV to HG-QBC 26.2 V Isc (short-circuit current) 8.21 A Voc (open-circuit voltage) 32.9 V Duty cycle 0.5 Switching frequency 50 kHz Inductors L1 = 0.2 mH L2 = 0.5 mH Capacitors C1 = 9000 µF C2 = 2000 µF Parameters . Values . Pmax (maximum power) 200 W Voltage at MPP 26.2 V VPV to HG-QBC 26.2 V Isc (short-circuit current) 8.21 A Voc (open-circuit voltage) 32.9 V Duty cycle 0.5 Switching frequency 50 kHz Inductors L1 = 0.2 mH L2 = 0.5 mH Capacitors C1 = 9000 µF C2 = 2000 µF Open in new tab The I–V and P–V curves of the SPV model for various irradiation levels are indicated in Fig. 14a and b. It is noted that the Vout and Pout are increased with an increase in the irradiation levels of 500, 750 and 1000 W/m2. Fig. 14: Open in new tabDownload slide I–V and P–V curves of SPV model at various irradiation levels. Fig. 15 shows the Vout of the SPV model. It is obtained as 26.2 V and is settled at 0.4 s. The Vout of the SPV module is quite low and it has to be enhanced for load integrated operation. The output of the CBC is twice the value of B, which results in a total output of 52.4 V with more ripple content. This voltage is not sufficient for load operation. To overcome the demerits of the CBC, the HG-QBC has been proposed. The value of B for the HG-QBC is 4. The Vout is four times the input voltage, or 104.8 V. Due to higher voltage gain, the CBC is replaced with the HG-QBC. Fig. 15: Open in new tabDownload slide Output voltage of SPV model. Fig. 16a and b shows the Vout and Pout curves of conventional BCs with the P&O-based MPPT at a value of G = 1000 W/m2. The Vout and Pout curves show more oscillations at 4.2%, a rise time of 0.18 s and time of settling of 0.2 s at the output. The value of B for the CBC is limited to 2, which leads to doubling the SPV panel voltage from 26.2 to 52.4 V. Fig. 16: Open in new tabDownload slide BC with traditional P&O-based MPPT for constant value of irradiation (a) output voltage and (b) power curves. Fig. 17a and b shows the Vout and Pout curves of the HG-QBC with traditional P&O-based MPPT at a constant value of G = 1000 W/m2. The simulated outcomes show that the Vout of the suggested HG-QBC provides the value of B about four times larger, which results in 102.7 V at a settling time of 0.5 s. The Pout is 173 W. Because of the slow convergence rate, higher output ripple, fluctuations at MPP and lower steady-state response, the CBC is replaced with the P&O-based MPPT for the HG-QBC. Fig. 17: Open in new tabDownload slide (a) Output voltage and (b) power curves for HG-QBC with traditional P&O-based MPPT for constant irradiation. The P&O-based MPPT is not suitable for rapidly varying environmental conditions because it fails to identify the fast and accurate MPP. Hence, the NN-based MPPT for the HG-QBC is proposed. At constant irradiation G = 1000 W/m2 the corresponding output voltage and power curves are indicated in Fig. 18a and b. The simulated results show that the Vout is 103.8 V and the power is 185.8 W at a settling time of 2 s. The NN-based MPPT controller technique tracks the MPP quite quickly and precisely. The performance is better with lower oscillations in association with traditional P&O-based MPPT for the HG-QBC. Fig. 18: Open in new tabDownload slide HG-QBC with NN-based MPPT for G = 1000 W/m2 for (a) output voltage and (b) power curves. The performance of the SPV output can be further enhanced. As a result, for an HG-QBC, a hybrid-based MPPT with NN assisted by the P&O technique is proposed. The Vout and Pout curves for the hybrid-based MPPT with NN assisted by P&O for an HG-QBC is depicted in Fig. 19a. The simulated results show that oscillations are reduced substantially in Vout and Pout curves with the use of hybrid-based MPPT with NN assisted by P&O for the HG-QBC. The voltage is 104.8 V and the power is 190.63 W. It converges faster with very low ripples at output, and with a settling period of 1.2 s, it quickly achieves a steady state compared to NN-based MPPT. Fig. 19: Open in new tabDownload slide HG-QBC with hybrid-based MPPT for constant irradiation curves for (a) output voltage and (b) power curves. To understand the performance characteristics of the SPV model under variable irradiation conditions with the proposed hybrid-based MPPT with NN assisted by P&O, it is simulated and discussed with the traditional P&O and NN-based MPPT for the HG-QBC. The simulation studies are carried out for different values of G. Fig. 20a and b shows the Vout and Pout curves for the HG-QBC with traditional P&O-based MPPT at various values of G. The simulation results show the highest Vout of 102.7 V and the highest Pout of 173 W for solar irradiations of 1000 W/m2. Fig. 20: Open in new tabDownload slide HG-QBC with traditional P&O-based MPPT for various values of G for (a) output voltage and (b) power curves. Fig. 21a and b shows the Vout and Pout curves for the HG-QBC with NN-based MPPT at different irradiation levels. The simulation results show the Vout of 103.8, 90.67 and 75.01 V and the output power of 185.8, 141.7 and 97.02 W for various values of G. The performance of the SPV model is improved in terms of Vout and Pout for the HG-QBC with NN-based MPPT at different irradiation levels. Fig. 21: Open in new tabDownload slide HG-QBC with NN-based MPPT for different irradiation curves for (a) output voltage and (b) power curves. The performance of the characteristic curves of the SPV model in terms of Vout and Pout for the HG-QBC with hybrid-based MPPT at different irradiation levels is depicted in Fig. 22a and b. The simulation results show the Vout of 104.8, 90.79 and 75.78 V and the Pout of 190.63, 143.1 and 99.69 W for different values of G. It is observed from Fig. 22a and b that, for different values of G, the oscillation is minimum in both the Vout and Pout curves with hybrid-based MPPT with NN assisted by P&O compared with the other two schemes of MPPT. Fig. 22: Open in new tabDownload slide HG-QBC with hybrid-based MPPT for different irradiation curves for (a) output voltage and (b) power curves. Table 3 indicates the comparison of various MPPT techniques with tracking efficiency. It can be decided that the hybrid-based MPPT technique has the highest tracking efficiency of the three MPPT methods. Table 3: Comparison of various MPPT techniques with tracking efficiency Type of MPPT . Speed . Precision . Efficacy . P&O Moderate Moderate 96% NN Fast Moderate—High 97% Hybrid Fast High 97.5% Type of MPPT . Speed . Precision . Efficacy . P&O Moderate Moderate 96% NN Fast Moderate—High 97% Hybrid Fast High 97.5% Open in new tab Table 3: Comparison of various MPPT techniques with tracking efficiency Type of MPPT . Speed . Precision . Efficacy . P&O Moderate Moderate 96% NN Fast Moderate—High 97% Hybrid Fast High 97.5% Type of MPPT . Speed . Precision . Efficacy . P&O Moderate Moderate 96% NN Fast Moderate—High 97% Hybrid Fast High 97.5% Open in new tab Analysis of several MPPTs in comparison for different BCs is tabulated in Table 4. The suggested hybrid-based MPPT with NN assisted by P&O has a settling time of 1.2 s and fewer voltage ripples. This suggests that the proposed MPPT is superior to other methods in terms of steady-state and dynamic response with fewer fluctuations at MPP. Table 4: Analysis of several MPPTs in comparison for different BCs Methods of MPPT . Type of converter . VPV(V) . Rise time (seconds) . Settling time (seconds) . Vout ripples (%) . Vout(V) . B . P&O Conventional 26.2 0.18 0.2 4.2 53.26 2.03 P&O HG-QBC 26.2 0.45 0.5 1.23 102.7 3.91 NN HG-QBC 26.2 1.8 2 0.61 103.8 3.96 Hybrid HG-QBC 26.2 1.08 1.2 0.20 104.8 4 Methods of MPPT . Type of converter . VPV(V) . Rise time (seconds) . Settling time (seconds) . Vout ripples (%) . Vout(V) . B . P&O Conventional 26.2 0.18 0.2 4.2 53.26 2.03 P&O HG-QBC 26.2 0.45 0.5 1.23 102.7 3.91 NN HG-QBC 26.2 1.8 2 0.61 103.8 3.96 Hybrid HG-QBC 26.2 1.08 1.2 0.20 104.8 4 Open in new tab Table 4: Analysis of several MPPTs in comparison for different BCs Methods of MPPT . Type of converter . VPV(V) . Rise time (seconds) . Settling time (seconds) . Vout ripples (%) . Vout(V) . B . P&O Conventional 26.2 0.18 0.2 4.2 53.26 2.03 P&O HG-QBC 26.2 0.45 0.5 1.23 102.7 3.91 NN HG-QBC 26.2 1.8 2 0.61 103.8 3.96 Hybrid HG-QBC 26.2 1.08 1.2 0.20 104.8 4 Methods of MPPT . Type of converter . VPV(V) . Rise time (seconds) . Settling time (seconds) . Vout ripples (%) . Vout(V) . B . P&O Conventional 26.2 0.18 0.2 4.2 53.26 2.03 P&O HG-QBC 26.2 0.45 0.5 1.23 102.7 3.91 NN HG-QBC 26.2 1.8 2 0.61 103.8 3.96 Hybrid HG-QBC 26.2 1.08 1.2 0.20 104.8 4 Open in new tab The detailed comparison of conventional P&O-, NN- and hybrid-based MPPT for the HG-QBC is tabulated in terms of voltage, current and power for different irradiation levels in Table 5. It is inferred that the proposed hybrid-based MPPT with NN assisted by P&O for the HG-QBC has the highest Vout at minimum oscillations. Table 5: Comparative analysis of various MPPT techniques under variable irradiations Irradiations (W/m2) . P&O . . . NN . . . Hybrid . . . . Voltage (V) . Current (A) . Power (W) . Voltage (V) . Current (A) . Power (W) . Voltage (V) . Current (A) . Power (W) . 500 74.46 1.221 90.88 75.01 1.293 97.02 75.78 1.316 99.69 750 90.17 1.478 133.33 90.67 1.563 141.7 90.79 1.578 143.1 1000 102.7 1.684 173 103.8 1.79 185.8 104.8 1.819 190.63 Irradiations (W/m2) . P&O . . . NN . . . Hybrid . . . . Voltage (V) . Current (A) . Power (W) . Voltage (V) . Current (A) . Power (W) . Voltage (V) . Current (A) . Power (W) . 500 74.46 1.221 90.88 75.01 1.293 97.02 75.78 1.316 99.69 750 90.17 1.478 133.33 90.67 1.563 141.7 90.79 1.578 143.1 1000 102.7 1.684 173 103.8 1.79 185.8 104.8 1.819 190.63 Open in new tab Table 5: Comparative analysis of various MPPT techniques under variable irradiations Irradiations (W/m2) . P&O . . . NN . . . Hybrid . . . . Voltage (V) . Current (A) . Power (W) . Voltage (V) . Current (A) . Power (W) . Voltage (V) . Current (A) . Power (W) . 500 74.46 1.221 90.88 75.01 1.293 97.02 75.78 1.316 99.69 750 90.17 1.478 133.33 90.67 1.563 141.7 90.79 1.578 143.1 1000 102.7 1.684 173 103.8 1.79 185.8 104.8 1.819 190.63 Irradiations (W/m2) . P&O . . . NN . . . Hybrid . . . . Voltage (V) . Current (A) . Power (W) . Voltage (V) . Current (A) . Power (W) . Voltage (V) . Current (A) . Power (W) . 500 74.46 1.221 90.88 75.01 1.293 97.02 75.78 1.316 99.69 750 90.17 1.478 133.33 90.67 1.563 141.7 90.79 1.578 143.1 1000 102.7 1.684 173 103.8 1.79 185.8 104.8 1.819 190.63 Open in new tab The suggested hybrid-based MPPT with NN assisted by P&O has a minimum settling time of 1.2 s with minimum Vout ripples (0.26%). This shows that the suggested MPPT technique performance is superior to the other two methods in terms of steady-state and dynamic response, as well as fewer oscillations in the MPP approach. 6 Conclusion At constant and different irradiation levels, the performance of the developed hybrid-based MPPT with NN assisted by P&O controller for the HG-QBC is compared to the conventional P&O- and NN-based MPPT for the SPV system. Using the MATLAB®/Simulink® platform, the SPV is modelled using a Kyocera 200GT-type module and the Vout and Pout curves are plotted under the constant and different values of G. The SPV is integrated with an HG-QBC for enhancing the output voltage at a higher level. The various existing BCs are compared in terms of the number of components count and voltage gain. The model of the HG-QBC is simulated in the MATLAB®/Simulink® environment, which provides a B of 4 and has high voltage gain for the SPV system in comparison with the CBC. The suggested hybrid-based MPPT is contrasted with a traditional P&O and the NN for the HG-QBC is presented. Furthermore, the simulation results of the hybrid-based MPPT are listed in Table 4 and it is noticed that the oscillations are minimum at MPP with negligible ripple, i.e. 0.20%, and a settling time of 1.2 s in comparison with the other two techniques. The performance of the HG-QBC with hybrid-based MPPT with NN assisted by P&O controller is better than the traditional P&O- and NN-based MPPT in terms of steady-state and dynamic responses, and at MPP it has lower oscillations. The proposed hybrid-based MPPT with NN and P&O controller converges faster and has better overall performance and efficiency. Conflict of interest statement None declared. References [1] Solis-Cisneros HI , Sevilla-Camacho PY, Robles-Ocampo JB, et al. A dynamic reconfiguration method based on neuro-fuzzy control algorithm for partially shaded PV arrays . Sustainable Energy Technol Assess , 2022 , 52 : 102147 . Google Scholar Crossref Search ADS WorldCat [2] Estévez-Bén AA , Alvarez-Diazcomas A, Rodríguez-Reséndiz J. Transformerless multilevel voltage-source inverter topology comparative study for PV systems . Energies , 2020 , 13 : 3261 . Google Scholar Crossref Search ADS WorldCat [3] Zuniga-Reyes MA , Robles-Ocampo JB, Sevilla-Camacho PY, et al. Photovoltaic failure detection based on string-inverter voltage and current signals . IEEE Access , 2021 , 9 : 39939 – 39954 . Google Scholar Crossref Search ADS WorldCat [4] Teh CJQ , Drieberg M, Soeung S, et al. Simple PV modeling under variable operating conditions . IEEE Access , 2021 , 9 : 96546 – 96558 . Google Scholar Crossref Search ADS WorldCat [5] Gonzalez-Diaz VR , Romero-Camacho S, Ambrosio-Lazaro RC, et al. A behavioral model for solar cells with transient irradiation and temperature assessment . IEEE Access , 2019 , 7 : 90882 – 90890 . Google Scholar Crossref Search ADS WorldCat [6] Singh O , Rajput SK. Mathematical modelling and simulation of solar photovoltaic array system . In: 2016 International Conference on Research Advances in Integrated Navigation Systems (RAINS) , Bangalore, India , 2016 , 1 – 5 . [7] Estévez-Bén AA , Alvarez-Diazcomas A, Macias-Bobadilla G, et al. Leakage current reduction in single-phase grid-connected inverters: a review . Applied Sciences , 2020 , 10 : 2384 . Google Scholar Crossref Search ADS WorldCat [8] Alvarez-Diazcomas A , López H, Carrillo-Serrano RV, et al. A novel integrated topology to interface electric vehicles and renewable energies with the grid . Energies , 2019 , 12 : 4091 . Google Scholar Crossref Search ADS WorldCat [9] López H , Rodríguez-Reséndiz J, Guo X, et al. Transformerless common-mode current-source inverter grid-connected for PV applications . IEEE Access , 2018 , 6 : 62944 – 62953 . Google Scholar Crossref Search ADS WorldCat [10] Ram JP , Babu TS, Rajasekar N. A comprehensive review on solar PV maximum power point tracking techniques . Renew Sustain Energy Rev , 2017 , 67 : 826 – 847 . Google Scholar Crossref Search ADS WorldCat [11] Ahmad J , Pervez I, Sarwar A, et al. Performance analysis and hardware-in-the-loop (HIL) validation of single switch high voltage gain DC-DC converters for MPP tracking in solar PV system . IEEE Access , 2020 , 9 : 48811 – 48830 . Google Scholar Crossref Search ADS WorldCat [12] Villegas-Mier CG , Rodriguez-Resendiz J, Álvarez-Alvarado JM, et al. Artificial neural networks in MPPT algorithms for optimization of photovoltaic power systems: a review . Micromachines , 2021 , 12 : 1260 . Google Scholar Crossref Search ADS PubMed WorldCat [13] Veerachary M , Shaw P. Controller design and analysis for fifth-order boost converter . IEEE Trans Ind Appl , 2018 , 54 : 4894 – 4907 . Google Scholar Crossref Search ADS WorldCat [14] Elsayad N , Moradisizkoohi H, Mohammed O. A new three-level flying-capacitor boost converter with an integrated LC2D output network for fuel-cell vehicles: analysis and design . Inventions , 2018 , 3 : 61 . Google Scholar Crossref Search ADS WorldCat [15] El-Helw HM , Magdy A, Marei MI. A hybrid maximum power point tracking technique for partially shaded photovoltaic arrays . IEEE Access , 2017 , 5 : 11900 – 11908 . Google Scholar Crossref Search ADS WorldCat [16] Ali A , Almutairi K, Padmanaban S, et al. Investigation of MPPT techniques under uniform and non-uniform solar irradiation condition: a retrospection . IEEE Access , 2020 , 8 : 127368 – 127392 . Google Scholar Crossref Search ADS WorldCat [17] Kiran Kumar BM , Indira MS, Nagaraja Rao S. Performance analysis of multiple gain boost converter with hybrid maximum power point tracker for solar PV connected to grid . Clean Energy , 2021 , 5 : 655 – 672 . Google Scholar Crossref Search ADS WorldCat [18] Zečević Z , Rolevski M. Neural network approach to MPPT control and irradiance estimation . Applied Sciences , 2020 , 10 : 5051 . Google Scholar Crossref Search ADS WorldCat [19] Hameed WI , Saleh AL, Sawadi BA, et al. Maximum power point tracking for photovoltaic system by using fuzzy neural network . Inventions , 2019 , 4 : 33 . Google Scholar Crossref Search ADS WorldCat [20] Essefi RM , Souissi M, Abdallah HH. Maximum power point tracking control using neural networks for stand-alone photovoltaic systems . International Journal of Modern Nonlinear Theory and Application , 2014 , 3 : 47805 . Google Scholar Crossref Search ADS WorldCat [21] Kumar BM , Rao SN, Indira MS. Analysis of grid-connected reduced switch MLI with high-gain interleaved boost converter and hybrid MPPT for solar PV . International Journal of Energy and Environmental Engineering , 2022 . doi:10.1007/s40095-022-00479-4. Google Scholar OpenURL Placeholder Text WorldCat [22] Ahmed HY , Abdel-Rahim O, Ali ZM. New high-gain transformer less DC/DC boost converter system . Electronics , 2022 , 11 : 734 . Google Scholar Crossref Search ADS WorldCat [23] Pirpoor S , Rahimpour S, Andi M, et al. A novel and high-gain switched-capacitor and switched-inductor-based DC/DC boost converter with low input current ripple and mitigated voltage stresses . IEEE Access , 2022 , 10 : 32782 – 32802 . Google Scholar Crossref Search ADS WorldCat [24] Anagha EB , Jos BM, Boby K, et al. Multiple output high gain DC-DC converter . Mater Today Proc , 2022 , 58 : 540 – 546 . Google Scholar OpenURL Placeholder Text WorldCat [25] Purushothaman SK , Sattianadan D, Vijayakumar K. Novel compact design high gain DC–DC step up converter applicable for PV applications . Sustainable Energy Technol Assess , 2022 , 51 : 101720 . Google Scholar Crossref Search ADS WorldCat [26] Banaei MR , Ardi H, Farakhor A. Analysis and implementation of a new single-switch buck-boost DC/DC converter . IET Power Electron , 2014 , 7 : 1906 – 1914 . Google Scholar Crossref Search ADS WorldCat [27] Shahir FM , Babaei E, Farsadi M. Voltage-lift technique based non-isolated boost DC–DC converter: analysis and design . IEEE Trans Power Electron , 2017 , 33 : 5917 – 5926 . Google Scholar Crossref Search ADS WorldCat [28] Chand MP , Ramesh G. Design of new positive output Super-Lift LUO Converter for solar input in comparison with different DC-DC converters . International Research Journal of Engineering and Technology , 2016 , 3 : 1588 – 1594 . Google Scholar OpenURL Placeholder Text WorldCat [29] Khanam JJ , Foo SY. Modeling of a photovoltaic array in MATLAB simulink and maximum power point tracking using neural network . J Electr Electron Syst , 2018 , 2 : 40 – 46 . Google Scholar OpenURL Placeholder Text WorldCat [30] Ali ZM , Alquthami T, Alkhalaf S, et al. Novel hybrid improved bat algorithm and fuzzy system based MPPT for photovoltaic under variable atmospheric conditions . Sustainable Energy Technol Assess , 2022 , 52 : 102156 . Google Scholar Crossref Search ADS WorldCat [31] Aleksandr L , Demidova G, Alassaf O, et al. MPPT algorithms for wind turbines: review and comparison . In: 2022 29th International Workshop on Electric Drives: Advances in Power Electronics for Electric Drives (IWED), Moscow, Russia , 2022, 1 – 6 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC [32] Basha CH , Murali M. A new design of transformer less, non-isolated, high step-up DC-DC converter with hybrid fuzzy logic MPPT controller . Int J Circuit Theory Appl , 2022 , 50 : 272 – 297 . Google Scholar Crossref Search ADS WorldCat © The Author(s) 2022. Published by Oxford University Press on behalf of National Institute of Clean-and-Low-Carbon Energy This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com © The Author(s) 2022. Published by Oxford University Press on behalf of National Institute of Clean-and-Low-Carbon Energy