Optimal selection of the decomposition structure based on GA for distributed model predictive control systems

A novel approach based on Genetic algorithm (GA) is proposed to find out the optimal decomposition structure for distributed model predictive control (DMPC) systems. The decomposition problem of DMPC requires a proper definition of the decomposition index and an efficient algorithm to solve the optimal decomposition optimization problem. In this paper, a new decomposition index representing both the coupling of subsystems and the communication load is defined. Besides, GA is utilized to solve the specific decomposition problem. To generate a new population for the improvement of the DMPC decomposition structure, the GA operators, including coding, selection, mutation and intercross of GA, are proposed to achieve the minimal coupling and the low communication load among subsystems. Finally, a ten-by-ten system is presented to demonstrate the effectiveness of the proposed algorithm.