Intelligent Control Systems using Computational Intelligence Techniques

Chapter 8: A Neural Network Approach for Nearly Optimal Control of Constrained Nonlinear Systems

Murad Abu-Khalaf, Frank L. Lewis

8.1 Introduction

The design of control systems requires one to consider various types of constraints and performance measures. Constraints encountered in control system design are due to physical limitations imposed on the controller and the plant. This includes actuator saturation, constraints on the states and others. Performance measures on the other hand are related to optimality issues. This includes minimum fuel, minimum energy, minimum time and other optimal control problems. In this chapter, a unified framework for constructing nearly optimal closed-loop controllers for constrained control systems is presented.

The control of systems with saturating actuators has been the focus of many researchers for many years. Several methods for deriving control laws considering the saturation phenomena are found in Saberi et al. [1] and Sussmann et al. [2]. Other methods that deal with constraints on the states of the system as well as the control inputs are found in Bitsoris and Gravalou [3], Hu et al. [4], Henrion et al. [5] and Gilbert and Tan [6]. This line of research is based on mathematical programming and the set invariance theory to design controllers that satisfy the required constraints. The controllers developed are not necessarily in closed-loop form. Moreover, optimality issues are not the main concern in this theme of work. Most of these methods do not consider finding optimal control laws for general nonlinear systems.

The optimal control of constrained input systems is well established. The controller can be found...

UNLIMITED FREE
ACCESS
TO THE WORLD'S BEST IDEAS

SUBMIT
Already a GlobalSpec user? Log in.

This is embarrasing...

An error occurred while processing the form. Please try again in a few minutes.

Customize Your GlobalSpec Experience

Category: Flow Controllers
Finish!
Privacy Policy

This is embarrasing...

An error occurred while processing the form. Please try again in a few minutes.