Decomposed Model Predictive Control for Economic Dispatch Problems

Model Predictive Control (MPC) was applied in Economic Dispatching (ED) problem, but its computation burden was large due to high complexity and dimension of problems. In this paper, a decomposed MPC approach is proposed to reduce the computational time without decreasing the performance of convergence and robustness. The decomposed MPC uses the "divide and conquer" concept, and divides the original problem into two less difficult sub-problems. The proposed MPC has been tested on a 6-unit system, and the results have been shown that decomposition can deliver promising solutions with less computation time.