论文部分内容阅读
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。
Aiming at the characteristics of instrument controller, this paper presents a practical closed-loop design method of instrument fuzzy neural network PID controller. The software of fuzzy neural network controller design and development is developed. A large number of computer simulations with different industrial background are made. Out of the controller parameters (network weight, network threshold) by the microcontroller control system. Experiments show that this method has a satisfactory curve effect on the output characteristics of the controlled object and the output characteristics of the D / A to the actuator. It also proves that this is an effective way to realize the intelligent instrumentation.