Neural Network PID Adaptive Control and Its Application

XU Chang,L(U) Jian-hong,CHENG Ming,ZHENG Yuan
DOI: https://doi.org/10.3969/j.issn.1671-7848.2007.03.016
2007-01-01
Abstract:A neural network PID control algorithm is presented to improve the PID adaptive control performance.Minimal resource allocating network(MRAN) that is an RBF network of self-tuning hidden nodes online and has the excellences of strong generation and high accuracy.The structure is improved by the reducing strategies of MRAN.The improved MRAN has more compact structure.Combining the improved MRAN with PID control,the PID self-adaptive control on dynamic RBF network is presented,and applied to control the superheated steam temperature of a power station boiler.Simulation results show that the presented algorithm has better performances than the conventional control algorithm.
What problem does this paper attempt to address?