Adaptive PID Controller Based on Single Neuron

摘要

An adaptive PID controller based on single neuron is proposed. The properties, control algorithm, parameters tuning, the control law and the application condition of the controller are studied in the paper. To satisfy the properties of the requirements of the control system in an electromotor group, such as a broad dynamic changing range, a fast response, a little overshoot and time-variable parameter, a new-type self-optimizing PID controller based on artificial neural networks is proposed and studied. It is verified that the controller has few adjustable parameters and excellent robust performance. The results of simulation and experiment prove that the controller is superior to the traditional PID controller.