基于反向云自适应粒子群算法的多目标无功优化

针对粒子群算法在高维复杂问题寻优时易陷入局部寻优的现象,提出了反向云自适应粒子群算法(OCAPSO)，通过反向学习加快算法的收敛速度，使用云模型来平衡粒子的全部搜索和局部搜索能力，使用自适应突变机制增强种群的多样性。用高维广义Schwarz函数对OCAPSO的有效性进行验证，进一步以IEEE30节点系统进行单目标和多目标无功优化测试并将测试结果与粒子群优化（PSO），进化算法（EA）等测试结果进行比较，证实了该算法的优越性。分析表明，OCAPSO算法用于解决多目标无功优化问题有效可行。

Multi-Objective Reactive Power Optimization Based on Opposition-based Learning Cloud Model Adaptive Particle Swarm Optimization

An opposition-based cloud model adaptive particle swarm optimization algorithm (OCAPSO) is presented to solve the high-dimensional problems that the conventional PSO algorithm is easy to fall into a locally optimized point. The algorithm convergence speed is accelerated through opposition-based learning, and the cloud model is used to balance the global and local search ability of each particle, and the adaptive mutation mechanism is used to enhance the population diversity. The effectiveness of OCAPSO is verified by high-dimensional generalized Schwarz function. Then single objective and multi-objective reactive power optimization of IEEE30 bus system are tested. The superiority of OCAPSO is confirmed by comparing with the testing results of PSO and EA. Analysis shows that OCAPSO is effective for multi-objective reactive power optimization.