联合紧支径向基函数和微分进化算法的电磁优化方法

Optimization design of electromagnetic devices based on combination of compactly supported radial basis function and differential evolution algorithm

针对工程电磁装置的优化设计问题,提出一种基于紧支径向基函数的近似模型和改进微分进化算法的组合优化方法。利用基于紧支径向基函数的近似模型得到多维设计空间中构造复杂目标函数和约束函数的显式函数关系。结合改进的微分进化算法获得精确度较高的全局最优解。实例分析和比较表明新方法的收敛速度和优化效率相对于其他的随机类优化算法和模型有明显的优势。

Aiming at engineering optimal design of electromagnetic devices, this paper proposes the optimization method based on a combination of compactly supported radial basis function approximation model and the improved differential evolution algorithm. Model the explicit function of complicated objective function and constraint for multi-dimensional design space struc- ture by Compactly Supported Radial Basis Function（CSRBF）. Obtain the high accuracy global optimal solution by the combina- tion of improved differential evolution algorithm. The analysis result shows that the new method of convergence rate and the optimize efficiency have obvious advantages compared to other random class.