PI-PD控制器的調整方法:利用模糊類神經網路

Abstract

在本篇論文，乃基於增益餘量與相位餘量的規格要求，利用模糊類神經網路提出一種調整PI-PD控制器的方法。首先，討論PI-PD控制器基於最佳化ISTE規格的調整方法和PI-PD控制器與PID控制器的差別。接下來基於增益餘量與相位餘量的規格要求調整PI-PD控制器的解析解，並發展適合此解析解的調整程序。由於此調整程序所得到的數學方程組太複雜，而模糊類神經網路在複雜的數學方程組上，可以得到一組合理又合於規格的解。所以，在本論文裡，利用模糊類神經網路，去得到所要求的規格。從模擬中，可知模糊類神經網路可以有效率地達到所要求的規格。

In this thesis, a method of tuning PI-PD controller with Fuzzy neural network is developed, which is based on specification of gain margin and phase margin. At first, discussion of tuning method of PI-PD controller based on ISTE specification is mentioned , then the difference between PI-PD controller and PID controller is stated. Next the analytic solution is achieved based on the specification of gain margin and phase margin, and the tuning process to this analytic solution is developed . Due to the complexity of mathematic equations in this tuning process, the Fuzzy neural network is applied to the complex equations for reasonable solutions. So the required specification can be obtained via Fuzzy neural network in this thesis. Simulation results show that the Fuzzy neural network can achieve the ideal specifications efficiently.