改进的禁忌搜索算法及其在电磁场逆问题中的应用

The application of improved tabu search algorithm in electromagnetic inverse problems

针对求解多极值点目标函数的电磁场逆问题,提出了一种改进的禁忌搜索算法。为提高算法的全局寻优能力和减少寻优计算的时间,改进算法在结构上提出了不同状态的新转移规则:步长向量选取原则和不同循环起始点选择的新规则。避免了重复或无意义中间点的产生,使算法能够有效地跳出局部极值点,迅速收敛到全局最优点。给出的典型数学函数验证了算法的快速性和有效性,电磁场逆问题算例TEAMWorkshop问题22的结果表明算法的优越性和工程应用价值。

An improved tabu search algorithm is presented to solve inverse electromagnetic problems for multimodal objective optimizations.In order to promote the global searching ability and decrease the calculation time, the algorithm is improved by a new transition rule for accepting new current states, i .e. a proper step vector choosing rule and a new rule for selecting initial states of different iterative cycles. These improvements avoid repeated mid-points, jump out of the local optimal and converge to the global optimal point efficiently .R esults of a typical mathematic function prove the rapidity and validity of the algorithm, and the results of item 22 of TEAM workshop which is an example of electromagnetic inverse problem are reported to show the excellent performance of the proposed algorithm and its great values for engineering application.