| Summary: | 碩士 === 義守大學 === 電機工程學系 === 92 === In this thesis, an auto-tuning PID-Like fuzzy controller is developed for adaptive control systems. In order to enhance the flexibility and control capability of the controller we developed, three parameters, including system error , error change and the change of error change , are used as the reference inputs of the fuzzy mechanism. Compare with the conventional fuzzy controller with two reference inputs (e and de), the controller we developed obviously could generate more appropriate force to control the system and then has better performances.
As we know, in many real dynamic systems, the mathematical model is unknown or the parameters of system are uncertain and may vary with time. For effectively controlling such a system, the controller must be designed to have enough adaptability. The relative parameters of controller can be efficiently auto-tuned in accordance with the immediate performance of the system. Therefore, the neural technique has been included to develop such a controller.
In our studies, an auto-tuning PID-Like fuzzy-neural controller is also studied and developed for adaptive control models, including periodic control and model reference control. For demonstrating the superiority of the controller we designed, several types of model reference control systems are studied and simulated. For a comparison, some types of fuzzy controller and auto-tuning fuzzy controller are also performed.
|
|---|
