Neural Network Control of a Class of Nonlinear Systems With Actuator Saturation

Abstract

A neural net (NN)-based actuator saturation compensation scheme for the nonlinear systems in Brunovsky canonical form is presented. The scheme that leads to stability, command following, and disturbance rejection is rigorously proved and verified using a general "pendulum type" and a robot manipulator dynamical systems. Online weights tuning law, the overall closed-loop system performance, and the boundedness of the NN weights are derived and guaranteed based on Lyapunov approach. The actuator saturation is assumed to be unknown and the saturation compensator is inserted into a feedforward path. Simulation results indicate that the proposed scheme can effectively compensate for the saturation nonlinearity in the presence of system uncertainty.

A robust anti-windup control design for electrically driven robots — Theory and experiment

“…A ProportionalDerivative (PD) controller in the actuated robotic system experiences this effect when magnitude of the control signal exceeds an acceptable actuator band. In this situation, the controller may accumulate errors and potentially cause unavoidable transient and steady-state deviations, instability, or even damaging the actuators [2,3]. To tackle this problem, many control strategies have been proposed aiming to prevent instability and nominal performance degradations.…”

“…The main discussion is that, all of three customary approaches for transmitted torque's measurement through motors shaft i.e., (I) torque measurement by using reaction force in shaft bearings, (II) the Prony brake method, and (III) torque measurement through induced strain in rotating body, suffer from several inherent weakness [12]. To overcome these weaknesses, a good work without saturated signal measurement was presented [3]. Besides, Ref.…”

“…[13] has proposed a static AW controller design for a passive planner robot manipulator without the gravitational term. However, [3] requires exact model of the input coefficients matrix, prior knowledge from nonlinear system dynamic, and an estimation of the bound of error. As a main weakness common in aforementioned strategies, almost previous approaches exclude the actuator dynamics.…”

Self-Structured Organizing Single-Input CMAC Control for De-icing Robot Manipulator

“…It is difficult to establish exactly mathematical model for the design of a model-based control system. In order to deal with this problem, the braches of current control theories are broad including classical control: neural networks (NNs) control [1][2][3], adaptive fuzzy logic control (FLCs) [4][5][6] or adaptive fuzzy-neural networks (FNNs) [7][8][9] etc. They are classified as adaptive intelligent control based on conventional adaptive control techniques where fuzzy systems or neural networks are utilized to approximate a nonlinear function of the dynamical systems.…”

“…Neural networks (NNs) are a model-free approach, which can approximate a nonlinear function to arbitrary accuracy [1][2][3]. However, the learning speed of the NNs is slow.…”