Disturbance rejection in setpoint tracking adaptive predictive controller design

The paper presents a new adaptive predictive control algorithm as well as the ability of disturbance rejection. The method combines together the concepts of system identification and adaptive predictive controller design. A multi-step output predictor is generated to derive the identified matrix by a standard recursive least-square technique. A control force is also calculated in term of input/output time histories. The formulation satisfies simultaneously system identification and adaptive predictive controller design requirement for tracking purposes and showing an excellent result in disturbance rejection.