Adaptive fuzzy control for high performance induction motor drives

Induction motor presents an extremely challenging control problem. It is rugged, simple and economic drive, however, it has complicated dynamics, their model is highly nonlinear and two of the state variables are usually unavailable for measurement. Moreover, the two control dynamics of the speed and the flux magnitude are coupled. Conventional control schemes introduced different techniques to decouple the control dynamics. However, these techniques are heavily model dependent and include a great deal of mathematics.; In the thesis, newly developed fuzzy control algorithms are introduced to overcome the difficulties encountered in the induction motor control system. The newly developed control algorithms have superior control performance and fast adaptation mechanisms. They enhance the control system accuracy and reduce the steady state error.; The thesis also introduces a multi-input/multi-output fuzzy model reference learning control that produces independent control for both the speed and the flux magnitude. The control algorithm is able to form the control surface that maps the actuating signals to the control actions. It also has the ability to treat the machine parameter variations by on-line adaptation of the control surface.; Finally, a new systematic procedure to design an adaptive fuzzy controller using genetic algorithms is introduced. The optimization process reduces the effort of tuning the parameters of a learning and adaptive fuzzy controller and ensures global optimization of these parameters.; This thesis presents the simulation and experimental results of the newly developed control algorithms applied to induction motor control system. The results show enhanced control performance with high adaptation and learning capabilities.