DISTRIBUTED MODEL PREDICTIVE CONTROL ALGORITHM BASED ON HIERARCHY DECOMPOSITION

摘要

Taking advantages of the connection structures of subsystems, the problem how to reduce communication burden in cooperative distributed model predictive control (DMPC) algorithms is researched in this paper, and the DMPC algorithm based on hierarchy decomposition is proposed. Using interpretative structural model (ISM) method, subsystems are divided into serial connected sets. During the iteration, subsystems only communicate with subsystems inside their connected sets, and do not communicate with subsystems outside their connected sets. The optimal inputs of subsystems are only sent to the downstream connected sets according to their sequences. A simulation example is done to show the efficiency of the proposed algorithm.