A fuzzy-neural-network-based visual feedback learning control for robot manipulators

To solve the problem of the delay, nonlinearity and time-varying properties of PMA-actuated knee-joint rehabilitation training device, a self-learning control method based on fuzzy neural network is proposed in this paper. A self-learning controller was designed based on the combination of pid controller, feedforward controller, fuzzy neural network controller, and learning mechanism. It was applied to the isokinetic continuous passive motion control of the PMA-actuated knee-joint rehabilitation training device. The experiments proved that the self-learning controller has the properties of high control accuracy and unti-disturbance capability, comparing with pid controller. This control method provides the beneficial reference for improving the control performance of such system.

Download Full-text

A Robust Adaptive Control using Fuzzy Neural Network for Robot Manipulators with Dead-Zone

International Journal of Computers Communications & Control ◽

10.15837/ijccc.2019.5.3621 ◽

2019 ◽

Vol 14 (5) ◽

pp. 692

Author(s):

D. Ha Vu ◽

Shoudao Huang ◽

T. Diep Tran ◽

T. Yen Vu ◽

V. Cuong Pham

Keyword(s):

Neural Network ◽

Fuzzy Neural Network ◽

Robot Manipulators ◽

Industrial Robot ◽

Dead Zone ◽

Approximation Error ◽

Robust Adaptive Control ◽

Neural Network Controller ◽

Fuzzy Neural ◽

Robust Adaptive

In this paper, a robust-adaptive-fuzzy-neural-network controller (RAFNNs) bases on dead zone compensator for industrial robot manipulators (RM) is proposed to dead the unknown model and external disturbance. Here, the unknown dynamics of the robot system is deal by using fuzzy neural network to approximate the unknown dynamics. The online training laws and estimation of the dead-zone are determined by Lyapunov stability theory and the approximation theory. In this proposal, the robust sliding-mode-control (SMC) is constructed to optimize parameter vectors, solve the approximation error and higher order terms. Therefore, the stability, robustness, and desired tracking performance of RAFNNs for RM are guaranteed. The simulations and experiments performed on three-link RM are provided in comparison with neural-network (NNs) and proportional-integral-derivative (PID) to demonstrate the robustness and effectiveness of the RAFNNs.

Download Full-text

Backstepping Adaptive Iterative Learning Control for Robotic Systems

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.284-287.1759 ◽

2013 ◽

Vol 284-287 ◽

pp. 1759-1763

Author(s):

Ying Chung Wang ◽

Chiang Ju Chien ◽

Chi Nan Chuang

Keyword(s):

Neural Network ◽

Iterative Learning Control ◽

Fuzzy Neural Network ◽

Tracking Error ◽

Learning Control ◽

Iterative Learning ◽

Robotic Systems ◽

Adaptive Iterative Learning Control ◽

Repetitive Tasks ◽

Fuzzy Neural

A backstepping adaptive iterative learning control for robotic systems with repetitive tasks is proposed in this paper. The backstepping-like procedure is introduced to design the AILC. A fuzzy neural network is applied for compensation of the unknown certainty equivalent controller. Using a Lyapunov like analysis, we show that the adjustable parameters and internal signals remain bounded, the tracking error will asymptotically converge to zero as iteration goes to infinity.

Download Full-text