以衡丰发电厂钢球磨煤机辨识结果为研究对象，基于MATLAB/Simulink平台搭建自适应神经网络模糊PID的磨煤机控制系统仿真模型,通过自适应神经网络对模糊规则进行训练和学习，改进磨煤机出口温度控制、入口负压控制以及负荷控制策略。仿真结果表明：自适应神经网络模糊PID控制优化效果明显，在磨煤机50%工况下，相比传统PID调节和模糊PID调节系统，稳定性分别提高57.96%和33.70%；调节速度分别提高43.88%和31.38%；稳态误差分别减少95.41%和89.33%。

Taking the identification results of the steel ball coal mill in Hengfeng Power Plant as the research object, a simulation model of the coal mill control system of the adaptive neural network fuzzy PID was built based on the MATLAB/Simulink platform, and the fuzzy rules were trained and learned through the adaptive neural network to improve the coal mill outlet temperature control, inlet negative pressure control and load control strategies.The simulation results show that the adaptive neural network fuzzy PID control optimization effect is obvious. Under 50% working conditions of the coal mill, compared with the traditional PID regulation and fuzzy PID regulation systems, the stabilities are increased by 57.96% and 33.70% respectively; the adjustment speeds are increased by 43.88% and 31.38% respectively; the steadystate errors are reduced by 95.41% and 89.33% respectively.

TP273+.4