Adaptive Control of Steam Generator Feed-Water System Based on Improved Single Neuron Self-Adaptive Control

Abstract

Because of the swell-and-shrink phenomena, the false water level of steam generator produces, and then it is difficult to control the water level of the steam generator. Presently the three-element controller in cascade configuration with two PID controllers is widely used. It is demonstrated that this method is effective and fast responsive. However, the major problem is that in dynamic condition the parameters of two PID controllers cannot be determined easily. An improved single neuron adaptive PID control strategy, which adopts the proportional coefficient K exponentially adaptively according to the error, is presented. The simulation results show that the feed-water control system based on improved single neuron adaptive PID controller has faster response speed, shorter regulate time and smaller oscillation and improves the dynamic performance distinctly.