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

Learn More →

Hybrid FES-Exoskeleton Control for Walking Gait Correction

Hybrid FES-Exoskeleton Control for Walking Gait Correction DE GRUYTER Current Directions in Biomedical Engineering 2022;8(3): 09-12 Chenglin Lyu*, Bennet Holst, Felix Röhren, Bernhard Penzlin, Steffen Leonhardt, and Chuong Ngo Hybrid FES-Exoskeleton Control for Walking Gait Correction https://doi.org/10.1515/cdbme-2022-2003 Abstract: Walking gait correction is fundamental to restor- ing body motion function for patients. Functional Electrical Stimulation (FES) is regarded as a promising method and es- sential to Neurotherapy, and the exoskeleton is fast becoming a key instrument in rehabilitation. Applying FES or exoskele- ton alone, however, has its inherent disadvantages. Therefore, the hybrid exoskeleton combined with FES promoted in recent years overcomes the deficiency of more degree of freedom control. In this paper, a hybrid FES-Exoskeleton for walking gait control was first proposed and evaluated. With the Force Sensing Resistors (FSR) sensor, the exoskeleton actively as- sists in walking. Simultaneously, it also triggers the FES of the soleus, tibialis anterior, and gastrocnemius muscles for dorsil- flexion and plantar flexion. Later, three different control strate- Fig. 1: Lower limb hybrid FES-Exoskeleton with FSR sensor and gies are employed for the pulse-width controlled FES. Even- two channels electrodes. tually, an ILC with a PID controller is applied in the hybrid exoskeleton, following the best foot angle trajectory. joint. In this work, we utilized the FES on the lower leg for Keywords: Functional Electrical Stimulation (FES), Hybrid additional joint control (plantar flexion and dorsilflexion). Exoskeleton, Walking gait control, ILC control, Force Sensing A hybrid FES-Exoskeleton for walking gait control was Resistors (FSR). first proposed and evaluated. This paper attempts to show the feasibility of the FES application on the exoskeleton for ankle control. The hybrid control system combines the Force Sens- 1 Introduction ing Resistors (FSR), which could detect the gait phase of the patient, and angle encode, which could transfer the real-time Restoring body motion function, for example, sit-to-stand and degree as feedback for the FES system. Later, three different walking, after stroke or spinal cord injury (SCI) is regarded as control strategies are integrated and compared for walking cor- a remarkable goal of rehabilitation. Functional electrical stim- rection. By employing the ILC with PID control, the overall ulation is a promising method and fundamental to Neurother- angle RMSE can be controlled under 5.42 during walking. apy [1]. Recent research investigates the FES application for body motion rehabilitation on upper limbs or lower limbs [2]. Along with the exoskeleton, the hybrid exoskeleton combined with FES may overcome the limitation of the exoskeleton or 2 Methods FES alone. Previous work has established that the exoskeleton could supply additional power for lower-body movements like 2.1 Exoskeleton walking or sit-to-stand [3]. Most of the exoskeletons provide lower limb assistance, in which the motors are mounted on An active lower limb exoskeleton is applied in the hybrid sys- the hip and knee. It lacks one degree of freedom on the ankle tem. Four motors providing additional torque when walking are mounted on the knee and hip. The whole frame of the ex- oskeleton is designed to assist lower body motion. Based on *Corresponding author: Chenglin Lyu, Medical Information the previous work [4], FSR sensors combined with angle en- Technology, Helmholtz-Institute for Biomedical Engineering, code, which are installed in the shoes, trigger the different legs RWTH Aachen University, Pauwelsstr.20, Aachen, Germany, for further FES control. e-mail: lyu@hia.rwth-aachen.de Open Access. © 2022 The Author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 International License 9 F. Author et al., Short title The hybrid exoskeleton is presented in Fig. 1. Stance leg and swing leg shift alternately during the gait. Via custom- designed thin-film FSR, the force data from the sole are mea- sured, from where the exoskeleton is triggered by one side of the FSR. The distinction between these two states is based on implemented threshold values and determines whether the test person is ready for the next step. Based on the motion trajecto- ries of hips and knees, a subordinate cascaded position control strategy is implanted on electric joint actuators. The design of the low-level controller assumes a gait cycle of about 2.8 s with a standstill pause of 0.1 s. The flowchart in Fig. 2 serves as an overview of the FSR-triggered gait phase control. Fig. 3: Angle-Pulse width relation for (l) dorsilflexion and (ll) plan- tar flexion of subject. Fig. 2: Flow chart for the FSR-triggered gait phase control. hand, a higher current covers more pulse width but will make the leg muscle group fatigue sooner, which is not suitable for long-term walking gait rehabilitation. On the other hand, a lower current could prevent fatigue to a certain extent, but it is hard to control. Therefore, the current and frequency of FES 2.2 FES are constant with 20 mA 30Hz for dorsilflexion and 25 mA 30 Hz for plantar flexion respectively. RehaMove3 stimulator (Hasomed GmbH, Magdeburg, Ger- many) is utilized to combine the hybrid exoskeleton. It can generate rectangular, bi-phasic-shaped impulses. These im- 2.3 Hybrid Control pulses provide current amplitude from 0 to 150 mA (resolution of 0.5 mA), pulse width from 10 to 4000 𝜇s, and frequency FES and exoskeleton are combined and synchronized to the from 1 to 500 Hz. Two pairs of electrodes are attached to the computer separately via the dSPACE and NI data acquisition. front and back of the leg to stimulate the soleus, tibialis ante- Fig. 4 shows the connection between the two systems, and also rior, and gastrocnemius muscles for dorsilflexion and plantar illustrates the hardware-in-the-loop test environment for the flexion control. exoskeleton including the FSR soles and the stimulator. FSR The angle of the footplate ranges from -12 (dorsal) to is not only part of the exoskeleton connected to the dSPACE, +18 (plantar). FES-induced muscle contractions are depen- but also connected to the DAQ system for synchronization and dent on three parameters, current, frequency, and pulse width. triggering. Pulse-width control is applied on FES, and the angle-pulse Principally, both systems run in parallel. The dSPACE box width relation between dorsal and plantar are evaluated. Fig. 3 is programmed in MATLAB/Simulink to control the exoskele- describes the angle and FES pulse width relation in dorsilflex- ton via the Control Desk. FES stimulator is operated in another ion and plantar flexion under the different currents (from 10 user interface to cooperate exoskeleton. Both models are trig- mA up to 40 mA) of 30 Hz frequency to find the optimal other gered by dSPACE´s Control Desk and an implemented phys- parameters in the FES application. In the experiments, the an- ical voltage change in the connection to the NI USB-6259 in- gle is detected by the angle encoder Orbis (RLS Merilna tech- tegrated into Simulink. This procedure provides an additional nika d.o.o., Zeje pri Komendi, Slovenia), which is a absolute safety feature and ensures the user to active stimulation for rotary encoder and consists of two components, a diametri- launch at the desired time. The signals from the FSR soles are cally magnetized permanent ring magnet and a printed circuit also transferred to the FES system via a branch of the cable. board. Except for the constrictions by FES, long-term muscle fa- tigue in FES control is needed to consider [5, 6]: On the one 10 F. Author et al., Short title Fig. 4: Hardware in the loop test environment for the exoskeleton and simulator. 3 Results The exoskeleton is evaluated with the trajectory following. As shown in Fig. 5, the left leg reference and output values of the knee motor M1 (I) and the hip motor (II) are given. The output of two motors reveals that the exoskeleton works well in the walking motion, and it is clear that the errors for the two joints in the control loop are relatively slight. Furthermore, phases 𝑎, 𝑏, 𝑐, 𝑑 represent the four walking Fig. 5: Evaluation of gait trial, exoskeleton control. (I) Reference postures from one sensor on the heel to the two sensors on the and actual angle of the knee motor M1, (II) reference and actual side of the middle foot and then to the one sensor on the toe. angle of the hip motor M2, (III) detected gait phases, (IV) deviation The chosen starting moment corresponds to the entry into the between detected gait phases between MATLAB and dSPACE. initialization phase 𝑒, which begins 1 s before the actual gait cycle. The gait phase-shifting detected by the threshold value tion control based on the well-performance exoskeleton for of the FSR is shown in the control panel (III). The gait phase the foot angle control. Table. 1 summarizes the three differ- detection implemented for dSPACE confirms that phases 𝑎 and ent strategies and their results in the following experiments in 𝑏 last correctly for 1.4 s, while phases 𝑐 and 𝑑 exit when the Fig. 6, where the comparison indicates that ILC & PID hybrid force conditions are satisfied. In addition, the deviation be- presents the best performance in gait control. tween the detected gait phases of Simulink and the control This experiment lasts 40 s long and records dorsilflexion panel is illustrated (IV). Both parts of the evaluation have been and plantar flexion. Angle trajectories reference and measure- preprocessed to begin at the same moment and display a 60 s ment data are contrasted (I), and then the RMSE of the angle gait experiment. per period over the entire experiment can be seen in Fig. 6-(II). The parameters of the ILC are inherited for the learning rate Tab. 1: Comparison of three different control strategies to be fixed at 𝛾 = 0.01 while the learn intervals Δ𝑡𝑙 = 0.3 s and 𝑃𝐼𝐷 = 0.025. In addition to the normalized ILC out- 𝑡ℎ𝑟𝑒𝑠 Duration Controller Parameters Angle RMSE put (III), both aligned PID outputs (IV) and the resulting pulse 1 40 s ILC Δ𝑡 = 0.3𝑠 6.24 widths (V) are also plotted. Further analysis shows that in gen- 2 40 s ILC Δ𝑡 = 0.4𝑠 6.93 eral ILC &PID control strategy follows the reference well and 3 40 s PID 𝑃𝐼𝐷 = 0.05 5.92 𝑡ℎ𝑟𝑒𝑠 decreases the overall angle RMSE around 5.42 . The reason 4 40 s PID 𝑃𝐼𝐷 = 0.025 5.46 𝑡ℎ𝑟𝑒𝑠 5 40 s ILC&PID Δ𝑡 = 0.3𝑠 & 5.42 why the RMSE fluctuates near the 5.42 could be the learning 𝑃𝐼𝐷 = 0.025 𝑡ℎ𝑟𝑒𝑠 time of ILC at the beginning. Besides, muscle fatigue would 6 40 s ILC&PID Δ𝑡 = 0.4𝑠 & 7.96 also be a factor that affects muscle construction. In summary, 𝑃𝐼𝐷 = 0.025 𝑡ℎ𝑟𝑒𝑠 the results of this study pave the way for a hybrid exoskeleton with an ILC and PID control on lower limb rehabilitation. If we now turn to the FES, we apply the ILC, PID, and PID & ILC combination in the functional electrical stimula- 11 F. Author et al., Short title the fatigue model of the lower leg and adjust dynamic ILC & PID parameters in control strategies. Author Statement The research was funded by the joint DFG-NSFC project "Hy- brid parallel compliant actuation for lower limb rehabilitation- HYPACAL" (LE817/34-1)(NSFC 51761135121). References [1] Cheung VCK, Niu CM, Li S, Xie Q, Lan N. A Novel FES Strategy for Poststroke Rehabilitation Based on the Natural Organization of Neuromuscular Control. IEEE Rev Biomed Eng. 2019;12:154-167. [2] I. L. Petersen, W. Nowakowska, C. Ulrich and L. N. S. A. Struijk, "A Novel sEMG Triggered FES-Hybrid Robotic Lower Limb Rehabilitation System for Stroke Patients," in IEEE Transactions on Medical Robotics and Bionics, vol. 2, no. 4, pp. 631-638, Nov. 2020. [3] Penzlin, B.; Bergmann, L.; Li, Y.; Ji, L.; Leonhardt, S.; Ngo, C. Design and First Operation of an Active Lower Limb Ex- oskeleton with Parallel Elastic Actuation. Actuators 2021, 10, [4] Penzlin B, Enes Fincan M, Li Y, et al. Design and analysis of a clutched parallel elastic actuator[C]//Actuators. Multidisci- plinary Digital Publishing Institute, 2019, 8(3): 67. [5] R. Riener, M. Ferrarin, E. E. Pavan and C. A. Frigo, "Patient- driven control of FES-supported standing up and sitting down: experimental results," in IEEE Transactions on Reha- bilitation Engineering, vol. 8, no. 4, pp. 523-529, Dec. 2000. [6] R. Riener and T. Fuhr, "Patient-driven control of FES- Fig. 6: ILC and PID control with Δ𝑡𝑙 = 0.3𝑠 and 𝑃𝐼𝐷 = 𝑡ℎ𝑟𝑒𝑠 supported standing up: a simulation study," in IEEE Trans- 0.025.(I) Reference and actual joint angle, (II) RMSE of the an- actions on Rehabilitation Engineering, vol. 6, no. 2, pp. 113- gle per period and overall, (III) ILC output, (IV) PID output, (IV) 124, June 1998. resulting pulsewidths. 4 Conclusion This study discussed the combination between FES and ex- oskeleton application in walking gait. The present research aimed to examine the control possibility of the hybrid ex- oskeleton, and the second aim of this study was to investigate different control strategies for walking gait correction. These experiments confirmed that FES can be used as rehabilitation and correction tools during walking, and PID with ILC control could better follow the ankle position. In general, therefore, it seems that FES-Exoskeleton could provide a new approach for drop foot and walking correction. Also, this combination of application and control strategies lay the groundwork for future research into lower body motion control. The question raised by this study is the precise angle control and muscle fa- tigue during the walking trails. Further research might explore http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Current Directions in Biomedical Engineering de Gruyter

Hybrid FES-Exoskeleton Control for Walking Gait Correction

Download PDF
4 pages

Loading next page...
 
/lp/de-gruyter/hybrid-fes-exoskeleton-control-for-walking-gait-correction-Jfpg5MLVzd
Publisher
de Gruyter
Copyright
© 2022 The Author(s), published by De Gruyter
eISSN
2364-5504
DOI
10.1515/cdbme-2022-2003
Publisher site
See Article on Publisher Site

Abstract

DE GRUYTER Current Directions in Biomedical Engineering 2022;8(3): 09-12 Chenglin Lyu*, Bennet Holst, Felix Röhren, Bernhard Penzlin, Steffen Leonhardt, and Chuong Ngo Hybrid FES-Exoskeleton Control for Walking Gait Correction https://doi.org/10.1515/cdbme-2022-2003 Abstract: Walking gait correction is fundamental to restor- ing body motion function for patients. Functional Electrical Stimulation (FES) is regarded as a promising method and es- sential to Neurotherapy, and the exoskeleton is fast becoming a key instrument in rehabilitation. Applying FES or exoskele- ton alone, however, has its inherent disadvantages. Therefore, the hybrid exoskeleton combined with FES promoted in recent years overcomes the deficiency of more degree of freedom control. In this paper, a hybrid FES-Exoskeleton for walking gait control was first proposed and evaluated. With the Force Sensing Resistors (FSR) sensor, the exoskeleton actively as- sists in walking. Simultaneously, it also triggers the FES of the soleus, tibialis anterior, and gastrocnemius muscles for dorsil- flexion and plantar flexion. Later, three different control strate- Fig. 1: Lower limb hybrid FES-Exoskeleton with FSR sensor and gies are employed for the pulse-width controlled FES. Even- two channels electrodes. tually, an ILC with a PID controller is applied in the hybrid exoskeleton, following the best foot angle trajectory. joint. In this work, we utilized the FES on the lower leg for Keywords: Functional Electrical Stimulation (FES), Hybrid additional joint control (plantar flexion and dorsilflexion). Exoskeleton, Walking gait control, ILC control, Force Sensing A hybrid FES-Exoskeleton for walking gait control was Resistors (FSR). first proposed and evaluated. This paper attempts to show the feasibility of the FES application on the exoskeleton for ankle control. The hybrid control system combines the Force Sens- 1 Introduction ing Resistors (FSR), which could detect the gait phase of the patient, and angle encode, which could transfer the real-time Restoring body motion function, for example, sit-to-stand and degree as feedback for the FES system. Later, three different walking, after stroke or spinal cord injury (SCI) is regarded as control strategies are integrated and compared for walking cor- a remarkable goal of rehabilitation. Functional electrical stim- rection. By employing the ILC with PID control, the overall ulation is a promising method and fundamental to Neurother- angle RMSE can be controlled under 5.42 during walking. apy [1]. Recent research investigates the FES application for body motion rehabilitation on upper limbs or lower limbs [2]. Along with the exoskeleton, the hybrid exoskeleton combined with FES may overcome the limitation of the exoskeleton or 2 Methods FES alone. Previous work has established that the exoskeleton could supply additional power for lower-body movements like 2.1 Exoskeleton walking or sit-to-stand [3]. Most of the exoskeletons provide lower limb assistance, in which the motors are mounted on An active lower limb exoskeleton is applied in the hybrid sys- the hip and knee. It lacks one degree of freedom on the ankle tem. Four motors providing additional torque when walking are mounted on the knee and hip. The whole frame of the ex- oskeleton is designed to assist lower body motion. Based on *Corresponding author: Chenglin Lyu, Medical Information the previous work [4], FSR sensors combined with angle en- Technology, Helmholtz-Institute for Biomedical Engineering, code, which are installed in the shoes, trigger the different legs RWTH Aachen University, Pauwelsstr.20, Aachen, Germany, for further FES control. e-mail: lyu@hia.rwth-aachen.de Open Access. © 2022 The Author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 International License 9 F. Author et al., Short title The hybrid exoskeleton is presented in Fig. 1. Stance leg and swing leg shift alternately during the gait. Via custom- designed thin-film FSR, the force data from the sole are mea- sured, from where the exoskeleton is triggered by one side of the FSR. The distinction between these two states is based on implemented threshold values and determines whether the test person is ready for the next step. Based on the motion trajecto- ries of hips and knees, a subordinate cascaded position control strategy is implanted on electric joint actuators. The design of the low-level controller assumes a gait cycle of about 2.8 s with a standstill pause of 0.1 s. The flowchart in Fig. 2 serves as an overview of the FSR-triggered gait phase control. Fig. 3: Angle-Pulse width relation for (l) dorsilflexion and (ll) plan- tar flexion of subject. Fig. 2: Flow chart for the FSR-triggered gait phase control. hand, a higher current covers more pulse width but will make the leg muscle group fatigue sooner, which is not suitable for long-term walking gait rehabilitation. On the other hand, a lower current could prevent fatigue to a certain extent, but it is hard to control. Therefore, the current and frequency of FES 2.2 FES are constant with 20 mA 30Hz for dorsilflexion and 25 mA 30 Hz for plantar flexion respectively. RehaMove3 stimulator (Hasomed GmbH, Magdeburg, Ger- many) is utilized to combine the hybrid exoskeleton. It can generate rectangular, bi-phasic-shaped impulses. These im- 2.3 Hybrid Control pulses provide current amplitude from 0 to 150 mA (resolution of 0.5 mA), pulse width from 10 to 4000 𝜇s, and frequency FES and exoskeleton are combined and synchronized to the from 1 to 500 Hz. Two pairs of electrodes are attached to the computer separately via the dSPACE and NI data acquisition. front and back of the leg to stimulate the soleus, tibialis ante- Fig. 4 shows the connection between the two systems, and also rior, and gastrocnemius muscles for dorsilflexion and plantar illustrates the hardware-in-the-loop test environment for the flexion control. exoskeleton including the FSR soles and the stimulator. FSR The angle of the footplate ranges from -12 (dorsal) to is not only part of the exoskeleton connected to the dSPACE, +18 (plantar). FES-induced muscle contractions are depen- but also connected to the DAQ system for synchronization and dent on three parameters, current, frequency, and pulse width. triggering. Pulse-width control is applied on FES, and the angle-pulse Principally, both systems run in parallel. The dSPACE box width relation between dorsal and plantar are evaluated. Fig. 3 is programmed in MATLAB/Simulink to control the exoskele- describes the angle and FES pulse width relation in dorsilflex- ton via the Control Desk. FES stimulator is operated in another ion and plantar flexion under the different currents (from 10 user interface to cooperate exoskeleton. Both models are trig- mA up to 40 mA) of 30 Hz frequency to find the optimal other gered by dSPACE´s Control Desk and an implemented phys- parameters in the FES application. In the experiments, the an- ical voltage change in the connection to the NI USB-6259 in- gle is detected by the angle encoder Orbis (RLS Merilna tech- tegrated into Simulink. This procedure provides an additional nika d.o.o., Zeje pri Komendi, Slovenia), which is a absolute safety feature and ensures the user to active stimulation for rotary encoder and consists of two components, a diametri- launch at the desired time. The signals from the FSR soles are cally magnetized permanent ring magnet and a printed circuit also transferred to the FES system via a branch of the cable. board. Except for the constrictions by FES, long-term muscle fa- tigue in FES control is needed to consider [5, 6]: On the one 10 F. Author et al., Short title Fig. 4: Hardware in the loop test environment for the exoskeleton and simulator. 3 Results The exoskeleton is evaluated with the trajectory following. As shown in Fig. 5, the left leg reference and output values of the knee motor M1 (I) and the hip motor (II) are given. The output of two motors reveals that the exoskeleton works well in the walking motion, and it is clear that the errors for the two joints in the control loop are relatively slight. Furthermore, phases 𝑎, 𝑏, 𝑐, 𝑑 represent the four walking Fig. 5: Evaluation of gait trial, exoskeleton control. (I) Reference postures from one sensor on the heel to the two sensors on the and actual angle of the knee motor M1, (II) reference and actual side of the middle foot and then to the one sensor on the toe. angle of the hip motor M2, (III) detected gait phases, (IV) deviation The chosen starting moment corresponds to the entry into the between detected gait phases between MATLAB and dSPACE. initialization phase 𝑒, which begins 1 s before the actual gait cycle. The gait phase-shifting detected by the threshold value tion control based on the well-performance exoskeleton for of the FSR is shown in the control panel (III). The gait phase the foot angle control. Table. 1 summarizes the three differ- detection implemented for dSPACE confirms that phases 𝑎 and ent strategies and their results in the following experiments in 𝑏 last correctly for 1.4 s, while phases 𝑐 and 𝑑 exit when the Fig. 6, where the comparison indicates that ILC & PID hybrid force conditions are satisfied. In addition, the deviation be- presents the best performance in gait control. tween the detected gait phases of Simulink and the control This experiment lasts 40 s long and records dorsilflexion panel is illustrated (IV). Both parts of the evaluation have been and plantar flexion. Angle trajectories reference and measure- preprocessed to begin at the same moment and display a 60 s ment data are contrasted (I), and then the RMSE of the angle gait experiment. per period over the entire experiment can be seen in Fig. 6-(II). The parameters of the ILC are inherited for the learning rate Tab. 1: Comparison of three different control strategies to be fixed at 𝛾 = 0.01 while the learn intervals Δ𝑡𝑙 = 0.3 s and 𝑃𝐼𝐷 = 0.025. In addition to the normalized ILC out- 𝑡ℎ𝑟𝑒𝑠 Duration Controller Parameters Angle RMSE put (III), both aligned PID outputs (IV) and the resulting pulse 1 40 s ILC Δ𝑡 = 0.3𝑠 6.24 widths (V) are also plotted. Further analysis shows that in gen- 2 40 s ILC Δ𝑡 = 0.4𝑠 6.93 eral ILC &PID control strategy follows the reference well and 3 40 s PID 𝑃𝐼𝐷 = 0.05 5.92 𝑡ℎ𝑟𝑒𝑠 decreases the overall angle RMSE around 5.42 . The reason 4 40 s PID 𝑃𝐼𝐷 = 0.025 5.46 𝑡ℎ𝑟𝑒𝑠 5 40 s ILC&PID Δ𝑡 = 0.3𝑠 & 5.42 why the RMSE fluctuates near the 5.42 could be the learning 𝑃𝐼𝐷 = 0.025 𝑡ℎ𝑟𝑒𝑠 time of ILC at the beginning. Besides, muscle fatigue would 6 40 s ILC&PID Δ𝑡 = 0.4𝑠 & 7.96 also be a factor that affects muscle construction. In summary, 𝑃𝐼𝐷 = 0.025 𝑡ℎ𝑟𝑒𝑠 the results of this study pave the way for a hybrid exoskeleton with an ILC and PID control on lower limb rehabilitation. If we now turn to the FES, we apply the ILC, PID, and PID & ILC combination in the functional electrical stimula- 11 F. Author et al., Short title the fatigue model of the lower leg and adjust dynamic ILC & PID parameters in control strategies. Author Statement The research was funded by the joint DFG-NSFC project "Hy- brid parallel compliant actuation for lower limb rehabilitation- HYPACAL" (LE817/34-1)(NSFC 51761135121). References [1] Cheung VCK, Niu CM, Li S, Xie Q, Lan N. A Novel FES Strategy for Poststroke Rehabilitation Based on the Natural Organization of Neuromuscular Control. IEEE Rev Biomed Eng. 2019;12:154-167. [2] I. L. Petersen, W. Nowakowska, C. Ulrich and L. N. S. A. Struijk, "A Novel sEMG Triggered FES-Hybrid Robotic Lower Limb Rehabilitation System for Stroke Patients," in IEEE Transactions on Medical Robotics and Bionics, vol. 2, no. 4, pp. 631-638, Nov. 2020. [3] Penzlin, B.; Bergmann, L.; Li, Y.; Ji, L.; Leonhardt, S.; Ngo, C. Design and First Operation of an Active Lower Limb Ex- oskeleton with Parallel Elastic Actuation. Actuators 2021, 10, [4] Penzlin B, Enes Fincan M, Li Y, et al. Design and analysis of a clutched parallel elastic actuator[C]//Actuators. Multidisci- plinary Digital Publishing Institute, 2019, 8(3): 67. [5] R. Riener, M. Ferrarin, E. E. Pavan and C. A. Frigo, "Patient- driven control of FES-supported standing up and sitting down: experimental results," in IEEE Transactions on Reha- bilitation Engineering, vol. 8, no. 4, pp. 523-529, Dec. 2000. [6] R. Riener and T. Fuhr, "Patient-driven control of FES- Fig. 6: ILC and PID control with Δ𝑡𝑙 = 0.3𝑠 and 𝑃𝐼𝐷 = 𝑡ℎ𝑟𝑒𝑠 supported standing up: a simulation study," in IEEE Trans- 0.025.(I) Reference and actual joint angle, (II) RMSE of the an- actions on Rehabilitation Engineering, vol. 6, no. 2, pp. 113- gle per period and overall, (III) ILC output, (IV) PID output, (IV) 124, June 1998. resulting pulsewidths. 4 Conclusion This study discussed the combination between FES and ex- oskeleton application in walking gait. The present research aimed to examine the control possibility of the hybrid ex- oskeleton, and the second aim of this study was to investigate different control strategies for walking gait correction. These experiments confirmed that FES can be used as rehabilitation and correction tools during walking, and PID with ILC control could better follow the ankle position. In general, therefore, it seems that FES-Exoskeleton could provide a new approach for drop foot and walking correction. Also, this combination of application and control strategies lay the groundwork for future research into lower body motion control. The question raised by this study is the precise angle control and muscle fa- tigue during the walking trails. Further research might explore

Journal

Current Directions in Biomedical Engineering – de Gruyter

Published: Sep 1, 2022

Keywords: Functional Electrical Stimulation (FES); Hybrid Exoskeleton,Walking gait control; ILC control; Force Sensing Resistors (FSR).

There are no references for this article.