A Methodology for Integrated System Identification and PID Controller Design

Abstract

Abstract This paper describes an integrated identification and control design methodology that begins with dynamic modeling from plant data and concludes with parameter settings for high performance PID controllers. By integrating identification with PID controller design, the method displays functionality that is often demanded by the practicing engineering community. The major steps in this integrated methodology are: experimental design and execution, high-order ARX estimation, and control-relevant model reduction leading to models that comply with the IMC-PID tuning rules. A high-order ARX model estimate serves as an intermediate model for control-relevant model reduction purposes; furthermore, the low computational effort associated with ARX estimation means that simple statistical tools (such as cross validation) can be used to efficiently determine a suitable structure for the ARX model without substantial user intervention. The methodology is demonstrated on a delayed time plant subjected to significant drift (i.e., a disturbance represented by an autoregressive integrated noise model).