A Novel Hybrid Multi-objective Optimization Algorithm and its Application to Designs of Eletromagnetic Devices

In this paper, a novel hybrid multi-objective optimization (MOO) algorithm is proposed by combining the improved Sparrow Search Algorithm (SSA) with the improved NSGA-II. The original SSA is improved by introduction of the population updating mechanism of Moth-flame Optimization (MFO) algorithm and by adopting adaptive mutation probability; meanwhile, the NSGA-II is enhanced by using Latin hypervolume sampling and dynamical selection mechanism of crossover and mutation operators. Numerical results of cases including TEAM22 benchmark problem and topology optimization problem of an electromagnetic actuator prototype demonstrate the effectiveness and superiority of the proposed hybrid algorithm.