Model reference adaptive control of an electromechanical servo system

The tuning of classical, servo system controllers is often problematical, owing to the presence of system nonlinearities and unmeasured disturbances. Adaptive control schemes are of growing interest since the system is self-tuned, such that some pre-defined performance requirement is met, regardless of variations in system parameters.

This paper describes the design and implementation of a microprocessor-based model reference adaptive controller for a small electromechanical ,servo system. The control scheme is based on an implementation of hyperstability theory, and robustness is enhanced by the incorporation of certain 'safety-jacket'features.

Experimental studies demonstrate that the controlled process tracks the reference model to a consistently high standard, even in the presence of perturbations to the process dynamics. It is also shown that the design of the error filter plays an important part in the speed of adaptation, and guidelines for error filter design are proposed.