| Summary: | 碩士 === 國立臺灣海洋大學 === 機械與機電工程學系 === 97 === Fuzzy controller is designed based on experience, in which the parameters of membership functions and fuzzy logic control rules are designed by trial and error, hence it is suitable for unknown mathematical model. When the more complex control system, the more importance of the fuzzy control. Therefore, if we can get the optimal parameters into the initial controller design, the fuzzy controller will be perfect. Genetic algorithm for the ability to search the optimal parameters and obtain the global optimal solution, and avoid a local rapidly.
In this paper, we proposed some modified genetic algorithms, such as elitist strategy, extinction strategy and immune algorithm, based on Visual Basic to design a Human Machine Interface of optimal parameters. Numerical simulation results show that the proposed controller approach has robustness and effectiveness.
|
|---|
