仪用模糊神经网络PID控制器的研究

The Study of Fuzzy Neural Network PID Controller for Instrument

针对仪用控制器的特点，本文提出了一种实用的仪用模糊神经网络ＰＩＤ控制器闭环设计方法，研制了模糊神经网络控制器设计开发软件，进行了大量的不同工业背景的计算机仿真，得出的控制器参数（网络权值，网络阀值）由单片机控制系统实现。实验表明：该方法在被控对象输出特性和Ｄ／Ａ输出给执行器的动作特性上都具有令人满意的曲线效果。同时亦证明这是一种实现仪器仪表智能化的有效方法。

According to the characteristic of controller for instrument, the paper proposes a closed loop design method of fuzzy neural network PID controller and developes a software on it.A large of computer stimulations on industrial models have also been made. The control algorithm is realized by a single chip microcomputer system. Experiments indicates that the method has both a satisfactory output of control system and output of D/A.It has also been proved that the method is an effective way to realize intelligence instrument.