Adaptive Neural Network Motion Control of Manipulators with Experimental Evaluations

Abstract

A nonlinear proportional-derivative controller plus adaptive neuronal network compensation is proposed. With the aim of estimating the desired torque, a two-layer neural network is used. Then, adaptation laws for the neural network weights are derived. Asymptotic convergence of the position and velocity tracking errors is proven, while the neural network weights are shown to be uniformly bounded. The proposed scheme has been experimentally validated in real time. These experimental evaluations were carried in two different mechanical systems: a horizontal two degrees-of-freedom robot and a vertical one degree-of-freedom arm which is affected by the gravitational force. In each one of the two experimental set-ups, the proposed scheme was implemented without and with adaptive neural network compensation. Experimental results confirmed the tracking accuracy of the proposed adaptive neural network-based controller.

Uncertainty estimation in robust tracking control of robot manipulators using the Fourier series expansion

“…Consider the problem of robust tracking control of electrically driven robot manipulators described by (14) and the control law (27). In order to estimate the lumped uncertainty F(t) using Fourier series, [T d1 , ..., T dn ] can be a good choice for the fundamental period duration of ξ i (t) in the Fourier series expansion (26).…”

“…If the estimated bound is greater than its actual value, the control system will face saturation of input and the chattering phenomenon, while too low estimation of the bounds may cause a higher tracking error. [12][13][14][15][16] The universal approximation property and linear parameterization are the most important motivations for these widespread applications. 11 Nowadays, various neural networks and fuzzy systems are extensively applied in the field of robot control.…”

“…[12][13][14][15][16] Conventional robust controllers 1-6 require nominal models, while the proposed control law is model-free. The proposed control law is model-free based on the voltage control strategy 18 (VCS) with the decentralized structure and only joint position feedback.…”

Adaptive back‐stepping cancer control using Legendre polynomials

“…Thus, model-free and robust/adaptive controllers are preferred [16][17][18][19]. During the last two decades, various neuro-fuzzy control approaches have been applied to robust control of many uncertain non-linear systems [20][21][22][23][24]. Cancer control is not exception and widespread applications of neural networks and fuzzy systems in cancer immunotherapy and other fields of medical sciences have been presented.…”

Robust model-free control of a class of uncertain nonlinear systems using BELBIC: stability analysis and experimental validation

“…Due to the advances in computer technology, there has been an increasing trend toward intelligent control of nonlinear systems. Various intelligent controllers including neural networks [2][3][4][5][6], fuzzy systems [7][8][9][10], reinforcement learning [11,12] have been presented. The reason for this widespread applications of intelligent control may be its independency to nominal models.…”