Improving Distributed Model Predictive Control Performance Via Weight Optimization using PSO

AbstractIn recent years there has been much research performed in developing Distributed Model Predictive Control (DMPC) techniques which allow a Model Predictive Control (MPC) scheme to be distributed amongst a number of agents. By optimizing the weights in an MPC system, performance can be improved. In this paper, a PSO based weight optimization method for a DMPC system is developed and it is shown how DMPC performance can be optimized whilst constraining the number of iterations of the optimization algorithm.