基于模糊神经网络PID的复合控制策略  

Compound Control Strategy Based on Fuzzy-Neural Network PID

作　　者：韦薇薇 何同祥[1] Wei Weiwei;He Tongxiang(Department of Control and Computer Engineering,North China Electric Power University,Hebei,Baoding,071003,China)

机构地区：[1]华北电力大学控制与计算机工程学院,河北保定071003

出　　处：《仪器仪表用户》2021年第1期90-93,共4页Instrumentation

摘　　要：为了结合模糊控制容错力强和神经网络PID在线学习和调整的优点,提出了一种结合模糊控制与神经网络PID控制的复合控制方法,即分别设计模糊控制器和神经网络PID控制器后,再利用权重分配器对这两个控制器进行权重分配来控制被控对象。将该控制策略应用于某火电机组的二级过热器减温水流量系统控制,并在simulink仿真平台进行仿真,仿真实验结果表明:该复合控制策略较传统的模糊控制或神经网络PID控制的上升时间更短,调节时间和超调量更小,稳态性能更好。In order to combine the advantages of fuzzy control with strong fault tolerance and neural network PID online learning and adjustment,a compound control method combining fuzzy control and neural network PID control is proposed.That is,the fuzzy controller and neural network PID controller are designed separately,and the weights are reused.The distributor assigns weights to the two controllers to control the object.The control strategy is applied to the secondary superheater desuperheating water flow system control of a thermal power unit,and the simulation is carried out on the simulink simulation platform.The simulation experiment results show that compared with traditional fuzzy control or neural network PID control,this compound control strategy has shorter rise time,smaller adjustment time and overshoot,and better steady-state performance.

关 键 词：复合控制 模糊控制 神经网络PID控制 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]