ADAPTIVE MULTILAYER SELF-TUNING HIGH-PERFORMANCE TRACKING CONTROL FORDC BRUSHLESS MOTOR

Adaptive control for high performance drive systems has recently becom e an important subject of research. In applications such as robotics, actuation, and manipulation, the rotor of the electric motor should fo llow a pre-selected track at all time. The tracking accuracy should no t be affected by parameter uncertainties, unknown load variations, or sudden external disturbances. In this paper two schemes of adaptive co ntrol are developed and tested for a dc brushless motor. The first sch eme is a single-layer self-tuning controller based on the Generalized Minimum Variance theory. The second is a multi-layer adaptive controll er consisting of a self-tuning control layer and a supervisory control layer. The supervisory controller continuously monitors the status of the system parameters, the structure of the controller, and the motor performance. A laboratory setup is constructed to test the proposed m ethods. Laboratory results show that the multi-layer controller is cap able of achieving the tracking process with a high degree of accuracy, even in the presence of large and sudden disturbances.