Chapter 6 - Model predictive, adaptive, and neural net controllers

Abstract

This chapter briefly introduces general types of controllers, beginning with model predictive controllers. A very powerful control methodology is model predictive control (MPC) that was first developed for the chemical process industries and is now being applied to a rapidly expanding range of technologies. The basic idea of MPC is to use a model to predict the output response of a system at various points in time in the future, and based on this, control inputs are computed to yield the desired response. In model predictive control, optimal constraints can be included in the solution for the control. Unlike proportional (P), proportional-plus-integral (PI), and proportional-plus-integral-plus-derivative (PID) control, the MPC approach considered in the chapter extends to the case when the system to be controlled has multiple inputs and outputs. In indirect adaptive control, the coefficients or parameters of the system model are estimated by first using input-output data, and the estimates are then used to generate the control.