Using Directional Genetic Algorithms to Design A Fuzzy Sliding Mode Controller
| Autor: | Lin, Don-Sen, 林東生 |
|---|---|
| Rok vydání: | 1996 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 84 Optimal parameter searching plays an important role in the design ofoptimal control systems. Utilizing an effective algorithm to find thetrue optimal parameter values is the main task of this thesis. Recently, genetic algorithms (GAs) have been widely used as an alter-native optimization scheme to the traditional optimization methods forfinding global optimal solutions. This is because that GAs have the fo-llowing advantages: (1) GAs emulate natural genetic operators such as crossover, mutation, and natural selection. (2) GAs are parallel, global search techniques. (3) GAs use objective function information instead of derivatives or other auxiliary knowledge. (4) The transition rule of GAs is probabilistic instead of determinis- tic. In this thesis, a genetic algorithm is applied to find optimal param-eters for fuzzy controllers. In addition to the commonly used opera- tors, such as reproduction, crossover, and mutation, we introduce an e-volutionary direction operator to keep the correct search direction andthis enhance the searching efficiency. Also fuzzy controller is appliedto a sliding mode control scheme to improve the steady-state error whi-ch exists inherently in it. Moreover, it will perform well for nonlinearsystems when the fuzzy membership functions are tuned with GAs. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|