Tuning of PID controllers for a system with Parameter Variations:A Fuzzy Neural approach

• Main Authors: Weu-xu Li, 李威緒
• Other Authors: Ching-Cheng Teng
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/98750133577358745099

碩士 === 國立交通大學 === 電機與控制工程系 === 88 === In the thesis, we present a PID tuning method for uncertain processes using Fuzzy Neural Network. This method has the property of high-speed and high-accuracy. The designed controllers make the system, which could be affected by external interference to change the parameters of transfer function, stable and meet the specifications of time domain or frequency domain . This is also the design principle of QFT. In text, a fuzzy neural network approach is proposed to identify the relationship between the specifications of time or frequency domain and the PID controller parameters. Then, the FNN is used to automatically tune the PI or PID controller parameters for the specifications so that neither numerical nor graphical methods need to be used. From the simulation results, we can know that the FNN indeed can performance well.