A computationally efficient optimal design algorithm of electromagnetic devices using adaptive response surface method

A global optimization algorithm combining an adaptive response surface approximation of the objective function and experiment design strategy is presented. In the adaptive approximation, an optimal Latin hypercube sampling strategy based on multi-objective Pareto optimization is developed to obtain the sampling data in the design variable space, and multiquadric radial basis function is employed to construct the response surface. The proposed algorithm is applied to an analytic example and TEAM problem 22, and its robustness and computational efficiency are investigated.