Multi-objective Optimization of Electromagnetic Launch System Based on Improved RVEA-IGNG Algorithm

For analyzing the influences of discharge sequence of each pulsed power supply and discharge voltage on the performance of the electromagnetic launch system, intelligent optimization algorithms are carried out based on the simulation model of the electromagnetic launch system. Based on the maximum rail current, projectile exit velocity and energy conversion efficiency, the corresponding multi-objective optimization model is established, and the Pareto front (PF) is acquired in RVEA-iGNG algorithm. The algorithm combines an evolutionary optimization algorithm with the reference vector adaptive adjustment by growing neural gas network. It can handle the multi-objective optimization of the electromagnetic launch system with complex irregular PF and achieve fast convergence. In order to provide decision makers with more diversified solutions, this paper proposes an improved RVEA-iGNG algorithm where the archive maintenance strategy is improved. The simulation results not only show that the RVEA-iGNG algorithm can effectively solve the electromagnetic launch system multi-objective optimization problem, but also show that the improved archive maintenance strategy is significantly better than the archive in the original RVEA-iGNG in terms of enhancing the diversity of solutions.