Papers by ho pham huy anh
Hysteresis modelling and compensation for piezoelectric actuator using Jaya-BP neural network
Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, Sep 16, 2021
This paper proposes a new training algorithm using a hybrid Jaya-back propagation algorithm (call... more This paper proposes a new training algorithm using a hybrid Jaya-back propagation algorithm (called H-Jaya) to optimize the neural network weights, which is applied to identify the nonlinear hysteresis Piezoelectric actuator based on the experimental input-output data. The identified H-Jaya-neural model will be used to design an advanced feed-forward (FF) controller for compensating the hysteresis nonlinearity. Furthermore as to improve the tracking performance, a feed-forward-feedback control scheme is conducted. To evaluate the effectiveness of the proposed approach, firstly, it is tested through identifying the nonlinear hysteresis of Piezoelectric (PZT) actuator and compared with other meta-heuristic techniques, including differential evolution (DE), particle swarm optimization (PSO), and Jaya. Then, the accuracy of the hysteresis model-based compensator is evaluated under various control experiments using the piezoelectric actuator. The results of experiments executed on PZT actuator configured with a PZS001 from Thorlabs prove that the proposed approach obtains an excellent performance in hysteresis modeling and compensation.
Advances in intelligent systems research, 2015
This paper describes a novel walking gait generation algorithm based on inverse kinematics for a ... more This paper describes a novel walking gait generation algorithm based on inverse kinematics for a biped robot. The proposed algorithm uses the PSO (Particle Swarm Optimization) algorithm in order to find optimized values for the five walking algorithm parameters. The proposed experiment approach is tested on a small-sized humanoid robot in the DCSELAB, VNU-HCM. Simulation and experimental results reveal that using the PSO algorithm with an efficient fitness function can significantly reduce learning time. Moreover, a considerable fast walking speed is achieved.
Design&Implementation an Adaptive Takagi-Sugeno Fuzzy Neural Networks Controller for the 2-Links Pneumatic Artificial Muscle (PAM) Manipulator using in Elbow Rehabilitation
This paper presents the design, development and implementation of an adaptive Takagi-Sugeno fuzzy... more This paper presents the design, development and implementation of an adaptive Takagi-Sugeno fuzzy neural networks (A-FNN) controller suitable for real-time manipulator control applications. The unique feature of the A-FNN controller is that it has dynamic self-organizing structure, fast learning speed, good generalization and flexibility in learning. The proposed adaptive algorithm focuses on fast and efficiently optimizing weighting parameters of A-FNN
Adaptive sliding mode control with hysteresis compensation-based neuroevolution for motion tracking of piezoelectric actuator
Applied Soft Computing, 2022
Hysteresis compensation and adaptive control based evolutionary neural networks for piezoelectric actuator
International Journal of Intelligent Systems, Jun 9, 2021
This manuscript introduces a new adaptive inverse neural (AIN) control method applied to precisel... more This manuscript introduces a new adaptive inverse neural (AIN) control method applied to precisely track the piezoelectric (PZT) actuator displacement. First, a 3‐layer neural network optimized by the enhanced differential evolution technique which modifies a mutation scheme and provides suggestions for selecting mutant coefficient F, crossover coefficient CR, and population size NP, is used to identify the inverse nonlinearity hysteresis structure of the PZT actuator. Second, a feed‐forward control based on the identified model is proposed to compensate for the PZT hysteresis effect. Third, the Lyapunov stability principle is used to design and implement an adaptive law‐based neural sliding mode model plus the feed‐forward compensator to ensure that the whole PZT plant is operated in asymptotical stability. The experiment results demonstrate the proposed AIN controller proves superiority in comparison with other advanced control methods.
Journal of Intelligent and Robotic Systems, May 16, 2017
This paper proposes the novel adaptive neural network (ADNN) compliant force/position control alg... more This paper proposes the novel adaptive neural network (ADNN) compliant force/position control algorithm applied to a highly nonlinear serial pneumatic artificial muscle (PAM) robot as to improve its compliant force/position output performance. Based on the new adaptive neural ADNN model which is dynamically identified to adapt well all nonlinear features of the 2-axes serial PAM robot, a new hybrid adaptive neural ADNN-PID controller was initiatively implemented for compliant force/position controlling the serial PAM robot system used as an elbow and wrist rehabilitation robot which is subjected to not only the internal coupled-effects interactions but also the external end-effecter contact force variations (from 10[N] up to critical value 30 [N]). The experiment results have proved the feasibility of the new control approach compared with the optimal PID control approach. The novel proposed hybrid adaptive neural ADNN-PID compliant force/position controller successfully guides the upper limb of subject to follow the linear and circular trajectories under different variable end-effecter contact force levels.
Engineering Applications of Artificial Intelligence, 2020
The nonlinearities hysteresis of the piezoelectric-PZT actuator can greatly degrade in precise po... more The nonlinearities hysteresis of the piezoelectric-PZT actuator can greatly degrade in precise positioning applications. Therefore, modeling and identifying the hysteresis parameter of PZT actuator are still many challenges nowadays. In this paper, the hybrid adaptive differential evolution and Jaya algorithm (aDE-Jaya) is proposed to identify the Bouc-Wen hysteresis model of a piezoelectric actuator. In the aDE-Jaya algorithm, the improvement is focused on a hybrid mutant operator ''DE/rand/1'' and Jaya operator tried to balance between two contradictory aspects of their performance: exploration and exploitation and adaptive control parameters (mutant factor F, crossover rate CR, population size NP) to enhance the convergence efficiency. To prove the effectiveness and robustness of the proposed aDE-Jaya algorithm, it is tested on 8 benchmark functions and compared with other state-of-the-art optimizations. The comparison results show that aDE-Jaya has better performance in convergence rate and accuracy. After that, aDE-Jaya is applied to identify the Bouc-Wen hysteresis model based on experimental input-output data. The identified Bouc-Wen hysteresis resulted is used to design the feedforward controller to test accurate identification. As a consequent, the proposed aDE-Jaya algorithm can successfully identify the highly hysteretic nonlinearity of the piezoelectric actuator with perfect precision.
Robust adaptive control of nonlinear dynamic systems using hybrid sliding mode regressive neural learning technique
Engineering Computations, May 30, 2023
Purpose The proposed Sliding Mode Control-Global Regressive Neural Network (SMC-GRNN) algorithm i... more Purpose The proposed Sliding Mode Control-Global Regressive Neural Network (SMC-GRNN) algorithm is an integration of Global Regressive Neural Network (GRNN) and Sliding Mode Control (SMC). Through this integration, a novel structure of GRNN is designed to enable online and. This structure is then combined with SMC to develop a stable adaptive controller for a class of nonlinear multivariable uncertain dynamic systems.Design/methodology/approach In this study, a new hybrid (SMC-GRNN) control method is innovatively developed.Findings A novel structure of GRNN is designed that can be learned online and then be integrated with the SMC to develop a stable adaptive controller for a class of nonlinear uncertain systems. Furthermore, Lyapunov stability theory is utilized to ensure the hidden-output weighting values of SMC-GRNN adaptively updated in order to guarantee the stability of the closed-loop dynamic system. Eventually, two different numerical benchmark tests are employed to demonstrate the performance of the proposed controller.Originality/value A novel structure of GRNN is originally designed that can be learned online and then be integrated with the sliding mode SMC control to develop a stable adaptive controller for a class of nonlinear uncertain systems. Moreover, Lyapunov stability theory is innovatively utilized to ensure the hidden-output weighting values of SMC-GRNN adaptively updated in order to guarantee the stability of the closed-loop dynamic system.
Implementation of adaptive fuzzy sliding mode control for nonlinear uncertain serial pneumatic-artificial-muscle (PAM) robot system
This paper proposes a new advanced adaptive fuzzy sliding mode control (AAFSMC) approach and thei... more This paper proposes a new advanced adaptive fuzzy sliding mode control (AAFSMC) approach and their suitability for use in the control of a highly nonlinear 2-dof serial PAM robot. Stability proof of the stable convergence of the overall closed-loop serial PAM robot system under novel AAFSMC control is demonstrated based on Lyapunov stability principle. Simulation results prove that the proposed AAFSMC control algorithm, applied to a two degree of freedom nonlinear serial PAM robot, are successfully designed and implemented with better performance and robustness in comparison with previous SMC and fuzzy SMC approaches.
Adaptive Evolutionary Neural Network Gait Generation for Humanoid Robot Optimized with Modified Differential Evolution Algorithm
This paper introduces a novel approach for the biped robot gait generation which aims to control ... more This paper introduces a novel approach for the biped robot gait generation which aims to control humanoid robot to walk more naturally and stably on a flat platform. The dynamic biped gait generator created by the novel adaptive evolutionary neural model (AENM) that is optimally identified with the proposed modified differential evolution (MDE) optimization algorithm. The comparison results with genetic algorithm (GA) and particle swarm optimisation (PSO) demonstrated the effectiveness of proposed MDE method. The prototype small sized humanoid robot is used to test the performance of the proposed MDE algorithm and other algorithms. The comparison results demonstrate that the new proposed neural AENM model proves an effective approach for a robust and precise biped gait generation.
Applied Intelligence, Aug 20, 2020
This paper introduces a novel adaptive inverse multilayer T-S fuzzy controller (AIMFC) optimally ... more This paper introduces a novel adaptive inverse multilayer T-S fuzzy controller (AIMFC) optimally identified with an optimization soft computing algorithm available for a class of robust control applied in uncertain nonlinear SISO systems. The parameters of multilayer T-S fuzzy model are optimally identified by the differential evolution (DE) algorithm to create offline the inverse nonlinear plant with uncertain coefficients. Then, the adaptive fuzzy-based sliding mode surface is applied to ensure that the closed-loop system is asymptotically stable in which the stability is satisfied using Lyapunov stability concept. The control quality of the proposed AIMFC algorithm is compared with the three recent advanced control algorithms applied in the Spring-Mass-Damper (SMD) benchmark system. Simulation and experiment results with different control parameters show that the proposed algorithm is better than the inverse fuzzy controller and the conventional adaptive fuzzy controller comparatively applied in both SMD system and the coupled-liquid tank system with the performance index using the least mean squares (LMS) error, which is investigated to demonstrate the efficiency and the robustness of the proposed AIMFC control approach.
Soft Computing, May 12, 2020
The piezoelectric actuator has been receiving tremendous interest in the past decade, due to its ... more The piezoelectric actuator has been receiving tremendous interest in the past decade, due to its broad applications in areas of micro-robotics, neurosurgical robot, MEMS, exoskeleton, medical applications, and other applications. However, the hysteresis nonlinearity widely existing in smart materials yields undesirable responses, which make the hysteresis control problem even more challenging. Therefore, many studies based on artificial neural networks have been developed to cope with the hysteresis nonlinearity. However, the back-propagation algorithm which is popular in training a neural network model often performs local optima with stagnation and slow convergence speed. To overcome these drawbacks, this paper proposes a new training algorithm based on the Jaya algorithm to optimize the weights of the neural NARX model (called Jaya-NNARX). The performance and efficiency of the proposed method are tested on identifying two typical nonlinear benchmark test functions and are compared with those of a classical BP algorithm, particle swarm optimization algorithm, and differential evolution algorithm. Forwardly, the proposed Jaya-NNARX method is applied to identify the nonlinear hysteresis behavior of the piezoelectric actuator. The identification results demonstrate that the proposed algorithm can successfully identify the highly uncertain nonlinear system with perfect precision.
System Modeling Identification and Control of the Two-Link Pneumatic Artificial Muscle Manipulator Optimized with Genetic Algorithms
In this paper, the application of modified genetic algorithms (MGA) in the parameterization of th... more In this paper, the application of modified genetic algorithms (MGA) in the parameterization of the 2-link pneumatic artificial muscle (PAM) manipulator is investigated. The optimum search technique, MGA-based ID method, is used to identify the parameters of the prototype 2-link pneumatic artificial muscle (PAM) manipulator described by an ARX model in the presence of white noise and this result will
Robotics and Autonomous Systems, Oct 1, 2017
This paper proposes a novel control system combining adaptively feed-forward neural controller an... more This paper proposes a novel control system combining adaptively feed-forward neural controller and PID controller to control the joint-angle position of the SCARA parallel robot using the pneumatic artificial muscle (PAM) actuator. Firstly, the proposed inverse neural NARX (INN) model dynamically identifies all nonlinear features of the SCARA parallel PAM robot. Parameters of the inverse neural NARX model are optimized with the modified differential evolution (MDE) algorithm. Secondly, combining the inverse neural NARX model that provides a feed-forward control value from the desired joint position and the conventional PID controller applied to improve the precision and reject the steady state error in the joint position control. Finally, the new adaptive back-propagation (aBP) algorithm, based on Sugeno fuzzy system, proposed for online updating the weight values of the inverse neural NARX model as to adapt well to the disturbances and dynamic variations in its operation. Experimental tests confirmed the performance and merits of the proposed control scheme in comparison with the traditional control methods.
Robust control of uncertain nonlinear systems using adaptive regressive Neural-based deep learning technique
Expert Systems With Applications, Mar 1, 2023
Hybrid Super-Twisting Sliding Mode and FOC Scheme for Advanced PMSM Driving Control
Springer eBooks, 2022
Push Recovery Based on Fly-Wheel Dynamics Control Applied to Stable Biped Robot Walking
Springer eBooks, 2022
Adaptive Fuzzy Sliding Mode Controller for Ball and Plate System Optimizing by Advanced Jaya Algorithm
Springer eBooks, 2022
Novel Rewinding-Time Technique Applied for Enhancing Meta-heuristic Algorithms
Springer eBooks, 2022
International Journal of Fuzzy Systems, Dec 5, 2017
This paper proposes a new cascade training multilayer fuzzy logic for identifying forward model o... more This paper proposes a new cascade training multilayer fuzzy logic for identifying forward model of multiple-inputs multiple-outputs (MIMO) nonlinear double-coupled fluid tank system based on experiment platform. The novel multilayer fuzzy model consists of multiple MISO model; for each MISO model, it composes of multiple single fuzzy Takagi-Sugeno (T-S) models. The cascade training using optimization algorithms optimally trained multilayer fuzzy model one by one. All parameters of multilayer fuzzy model were optimally and comparatively identified using DE, GA and PSO optimization algorithms. Then, the proposed method results are compared with normal training method results. The experimental results show that proposed method gives better performance than the normal training. Hence, the novel proposed optimized multilayer fuzzy model is efficiently applied for identifying MISO system. The experiment cascade training is clearly presented. It proves more accurate and takes less time to compute than the normal training, and it seems promisingly scalable as a simple and efficient method to successfully identify and control various uncertain nonlinear large-scale MIMO systems.
Uploads
Papers by ho pham huy anh