Direct adaptive generalized predictive control. Application to motor drives with flexible modes

Predictive control has proved to be an efficient control strategy, and its application in the motor drives field has provided good results, even with small sampling rates. However, in case of important parameters variations, a high level of performance may be impossible to maintain with a fixed controller, so that an adaptive version becomes necessary. This paper presents a direct adaptive generalized predictive control structure, including a least squares-type strategy for the controller parameters on-line identification, and a conditional updating test. An application to motor drives with flexible transmissions is finally developed, taking into account some variations of inertia.