基于改进差异进化算法的有功优化仿真研究

Simulation Research on Active Power Optimization by Improved Differential Evolution Algorithm

在电力系统有功优化这个复杂的全局优化问题上,差异进化(Differential Evolution,DE)算法可以增加其种群多样性但搜索效率低,于是在其基础上提出了一种改进的差异进化算法(Improved Differential Evolution,IDE)。IDE算法保留了DE算法的三大步骤:变异、交叉以及选择,优化了传统的变异策略,同时引入了Logistic映射改变系统参数,使固定取值的搜索步长和交叉算子在一定范围内随机取值,以此扩大种群搜索范围,加快收敛速度;IDE算法最后运用了考虑系统约束的非贪婪选择,以确保算法在可行域里探索最优解。为验证算法的实用性,利用Matlab软件,将DE和IDE算法在IEEE30节点测试系统上进行目标函数为电网功率损耗的有功优化仿真。仿真结果表明,IDE算法增加了种群多样性,加快了收敛速度并且提高了搜索效率。通过此次仿真,加深了学生对电力系统有功优化以及DE算法的认识和理解,同时引导学生利用计算机技术改善算法的搜索性能并且求解优化问题。

In the complex global power system optimization, the differential evolution （DE） algorithm can increase the diversity of its population, but the search efficiency is low. Hence, an improved differential evolution （IDE） algorithm is proposed based on the DE algorithm. The IDE algorithm retains the three steps of DE algorithm ： mutation, crossover and selection, but it optimizes the traditional mutation strategy and combines with the Logistic mapping to make the search step and crossover change from fixed value to a certain range of random value. The method can expand the search scope of population and speed up the convergence rate. The IDE algorithm finally uses the non-greedy choice which considers system constraints to ensure that the algorithm explores the optimal solution in the feasible domain. In order to illustrate the practicability of the algorithm, by the Matlab software, DE algorithm and IDE algorithm were implemented on the IEEE30 bus test system for the active power optimization, and the objective function is power losses. Simulation results show that the IDE algorithm increases the population diversity, speeds up the convergence rate and improves the search efficiency. The simulation experiment deepens students＇ understanding of active optimization of power system and DE algorithm. At the same time, it can guide students to improve the search performance of algorithms and solve the optimization problems by using computer technology.