A Kriging-Assisted Light Beam Search Method for Multi-Objective Electromagnetic Inverse Problems

作者：An, Siguang; Yang, Shiyou^*; Mohammed, Osama A.

来源：IEEE Transactions on Magnetics, 2018, 54(3): 7001104.

DOI：10.1109/TMAG.2017.2748560

• Decision making
• inverse problems
• Pareto optimization
• surrogate modeling

A kriging-assisted light beam search (LBS) method is proposed to solve multi-objective inverse problems. To reduce the computational burden and increase the convergence speed, a kriging model is introduced into the evolutionary procedure of the LBS method. To guarantee the accuracy of the final Pareto solutions, a dynamic detecting strategy is used in the LBS method. To reflect the preference of a decision maker (DM) in decision making, a boundary control mechanism is proposed to assure that all the obtained Pareto solutions are well-distributed within the preference of the DM. To testify the accuracy of the proposed method, a typical test function, a benchmark inverse problem, TEAM Workshop Problem 22, and a linear antenna array design are solved. The numerical results demonstrate the effectiveness and efficiency of the proposed method.

• 出版日期2018-3
• 单位浙江大学; 中国计量大学