Comparison of Conventional and Repetitive MPC with Application to a Periodically Disturbed Compressor Stator Vane Flow

Abstract The paper explores and compares the closed-loop behavior of conventional Model Predictive Control (MPC) and Repetitive Model Predictive Control (RMPC) in the presence of unknown, highly transient disturbances which are periodic in time. Since MPC applies information about future disturbances when predicting the future response of the system to be controlled, the presence of unknown and transient disturbances, may decrease the quality of the prediction and thus also the performance of the MPC-controlled system significantly. If the disturbance is periodic in time, however, this problem can be attenuated. RMPC approaches the loss of prediction precision by integrating output measurements from past periods to the future forecast. The paper reviews both MPC algorithms for linear systems briefly and compares them experimentally to each other by means of regulating a compressor stator vane flow with transient, periodic disturbances while respecting the input and output constraints of the system.