Applications of adaptive differential evolution to optimize and identify the parameters of power electronics and electric machines

Abstract

Adaptive Differential Evolution is a derivative of Differential Evolution with adaptive differential weight and crossover rate, which has been evaluated by various benchmark functions. The Adaptive Differential Evolution inherits the merits of conventional Differential Evolution to find global optimal solutions with a faster and smoother convergence than the conventional heuristic algorithm, e.g., conventional Genetic Algorithm. The Adaptive Differential Evolution can find the global optimal solutions more steadily regarding numerous single-objective and multi-objective systems, while the conventional Genetic Algorithm owns the risk of finding local optimal solutions. In this Chapter, the Adaptive Differential Evolution algorithms are compared with conventional Genetic Algorithm in optimizing and identifying the parameters of power electronics systems and electric drives. Simulation and experimental results validate that the Adaptive Differential Evolution can reduce the operating cost of a direct-current microgrid, optimize torque, energy efficiency and torque ripple of an electrical continuously variable transmission system, identify the parameters of a series-series compensated wireless power transfer system, and identify the d-axis inductance, the q-axis inductance and the stator resistance of a dual-rotor flux modulated machine.