An Improved Random Taguchi's Method Based on Swarm Intelligence and Dynamic Reduced Rate for Electromagnetic Optimization

To avoid being easily trapped in a local optimal solution of Taguchi's method, an improved random Taguchi's method (IRTM) based on swarm intelligence and dynamic reduced rate (DRR) function is presented in this letter. The DRR is linked to the design parameter's percentage contribution toward performance variance obtained from the analysis of variance. To demonstrate the efficiency of the IRTM, four test function benchmarks and the design of an aperiodic planar array are investigated. Moreover, the results of the IRTM are compared with that of the classic meta-heuristic algorithms. From the comparison results, the IRTM is more efficient in that better performance could be obtained with less experiment times.