A Self-Tuning PID Control Method for Multi-Input-Multi-Output Nonlinear Systems

Abstract

In this study, an artificial neural network (ANN) model-based self-tuning PID control method is proposed for the control of multi-input-multi-output (MIMO) nonlinear systems. A single layer, feed-forward ANN structure is trained via input and output data randomly collected from the system and classified as learning, test, and validation data to obtain the system model. The obtained model is utilized in an adaptive PID control scheme in conjunction with two different optimization methods for PID tuning and control. Using this scheme, PID parameters can be tuned to their optimum values and the system can be controlled simultaneously. The performance of the proposed method is demonstrated via experimental studies.