Optimization design of separated dry-type air-core reactor based on modified differential evolution algorithm

According to the characteristics of nonlinear, multivariable and multiple constraint of the optimization of reactor, a modified differential evolution algorithm is proposed based on hybrid mutation and diversity-guided strategy. The algorithm carries out two forms of hybrid mutation in a given probability. The search of the individual is guided by the diversity information when carrying out the mutation, namely we change the search direction of the individual and make it escape the center point when the distance between the individual in differential vector or recombinant vector and the center point is less than the given lower limit. The simulation results of separated dry-type air-core reactor show that the HMDGDE algorithm is of stronger global search ability and better stability than the initial differential evolution (DE) algorithm and DE/rand/1/either-or algorithm.