The Use of Artificial Intelligence in Self-Tuning PID Controllers

摘要

In this paper a self-tuning PID controller, a neural network controller and a fuzzy expert systems are firstly compared and then combined into a hierarchical controller with a PID structure. The overall controller then makes better use of computational characteristics but retains a simple PID structure. The overall controller achieves an improvement in control performance for an unknown process operating in a changing environment or varying conditions of operation.