基于改进和声搜索算法的电网多目标差异化规划  

Multi-objective differential planning based on improved harmony search algorithm for power network

作　　者：宋春丽[1] 刘涤尘[1] 吴军[1] 王浩磊[1] 赵一婕[1] 潘旭东[1] 董飞飞[1] 

机构地区：[1]武汉大学电气工程学院,湖北武汉430072

出　　处：《电力自动化设备》2014年第11期142-148,共7页Electric Power Automation Equipment

基　　金：国家自然科学基金资助项目(51207114);国家电网公司大电网重大专项资助项目课题(SGCC-MPLG029-2012)~~

摘　　要：为提高电网的抗灾能力需对电网进行差异化规划。用全寿命周期内的新增成本和减损效益对规划的经济性和可靠性进行量化,建立了多目标差异化规划优化模型。根据Pareto支配关系制定寻优准则,以效益成本比最大为目标从Pareto解集中选择最优的差异化规划方案。将和声搜索算法应用于模型的求解,引入混沌映射、动态参数设置、改进的音高调整策略以及和声寻优信息共享机制使其能够解决0-1规划问题,并提高其搜索性能。对IEEE 30节点算例的分析表明,该模型正确合理,能得到最优规划方案。与常规的遗传算法和粒子群优化算法进行对比,改进和声搜索算法收敛速度快,具有较好的适应性和鲁棒性。The differential planning of power network should be carried out for enhancing its resilience to disasters. An optimal multi-objective differential planning model is established for power network,which applies the additional cost and the benefit of loss reduction covering whole life cycle to quantify its economy and reliability. The optimization criteria is set according to the Pareto dominate relation,which takes the maxi- mum benefit-cost ratio as its objective to select the optimal differential planning scheme from the Pareto solution set. The harmony search algorithm is applied to solve the model and new technologies are adopted to improve its search performance and solve the 0-1 planning problem,such as chaotic mapping,dynamic parameter setting,improved pitch adjustment strategy,harmony optimization information sharing,etc. Analysis of IEEE 30-bus case shows that,being correct and reasonable,the model can obtain the optimal planning scheme. Compared with the traditional genetic algorithm and the particle swarm optimization algorithm,the improved harmony search algorithm has better convergence speed,adaptability and robustness.

关 键 词：电力系统 规划 多目标差异化规划 成本 减损效益 改进和声搜索算法 可靠性 经济性 模型 

分 类 号：TM715[电气工程—电力系统及自动化]