Improved Multi-goal Particle Swarm Optimization Algorithm and Multi-output BP Network for Optimal Operation of Power System

To achieve the optimal operation of power system, an improved multi-goal particle swarm optimization (IMPSO) algorithm is proposed in this paper. Based on the multi-goal optimal power flow (MOOPF) calculation, IMPSO algorithm can determine high-quality scheduling schemes which effectively reduce fuel cost, power loss and exhaust emission. Compared with the basic multi-goal PSO (BMPSO) algorithm, IMPSO algorithm realizes better solution diversity and searching ability by integrating an innovative dominant strategy and the mutation-crossover operation of inferior solutions. Four experiments prove that the proposed IMPSO algorithm achieves more superior Pareto optimal set (POS) and best compromise scheme (BCS) than BMPSO algorithm. Furthermore, the multi-output BP power flow prediction model is put forward in this paper to seek the winning elite schemes (WES) around the BCS of IMPSO algorithm. The presented BP prediction model can find multiple WES schemes of bi-objective and tri-objective MOOPF problems, which realize zero constraint violation and smaller goals. In general, the superior WES schemes determined by proposed IMPSO and BP power flow prediction model are of great help to realize the optimal operation of power system with improved economy and safety.