改进EM算法在限流措施优化配置中的应用

Application of electromagnetism-like mechanism in short circuit current limiting optimization

仿电磁学算法(Electromagnetism-like mechanism,EM)是一种随机全局优化算法,它通过模拟电磁场的吸引和排斥机制,将每个解比作一个带电粒子,然后按一定准则使得搜索粒子朝最优解移动。本文对该算法进行改进,用均匀设计方法使初始种群更均匀,用混沌法优化局部搜索使其收敛加速,改进了力的计算和粒子的更新,最后将改进的EM算法用于限流措施优化配置中。此算法在求解含有离散变量和非线性约束条件的优化问题时具有实现方便、效率高的优点,使得限流措施优化有很好的收敛性和全局寻优能力。通过对新英格兰39节点系统进行计算及结果分析,说明所提出算法的全局搜索能力强、效率高,能得到较好的结果。

Electromagnetism-like mechanism（ EM） is a stochastic global optimization algorithm. It simulates the mechanism of attraction and repulsion in the electromagnetic field,assimilates each solution to a charged particle, and according to certain criteria makes the search particles move towards the optimal solution. In this paper,we improve the algorithm： using uniform design method makes the initial population more uniform; using chaotic optimization for local search makes convergence acceleraed; and calculation of force and particle update also makes some improvements. Finally the improved EM algorithm is used in the optimal configuration of current limiting measures. The algorithm for solving the discrete variable and nonlinear constraint optimization problem has the advantage of convenience and high efficiency,and it makes current limiting optimization to have good convergence and global optimization ability. The calculation and result analysis of New England 39 nodes system show that the proposed algorithm has better global searching capability,high efficiency,and can get good results.