Abstract: An “analytic+numerical” optimization method is proposed for the structural parameter optimization of magnetic shielding device, which combines the calculation simplicity of analytic method and the high calculation accuracy of numerical method. First, it is necessary to judge whether there is an analytical expression for the calculation of magnetic shielding performance. If it is difficult to obtain the analytical expression, the artificial neural network can be used to establish a calculation model, and a “black box analytical expression” is obtained. The intelligent optimization algorithm is used for quick “rough” optimization for analytical expression or “black box analytical expression” in a short time. And then the numerical method is used for the fine scanning calculation based on the “rough” optimal solution to achieve a “fine” optimization. The credibility of the optimized results is confirmed by cross-validation. The structural parameter optimization of double layer cylindrical and single layer rectangular magnetic shielding devices were simulated. The simulated results show that the coarse-to-fine optimization method is correct, and has higher reliability than pure analytical optimization. Key

Key words: magneticshielding, structuralparameteroptimization, combinatorialoptimization, cross-validation