梯级水库群优化调度精英集聚蛛群优化方法  

Multi-reservoir operation using elite-gather social spider optimization

作　　者：冯仲恺[1] 牛文静[1] 程春田[1] 武新宇[1] 

机构地区：[1]大连理工大学水电与水信息研究所,辽宁大连116024

出　　处：《水利学报》2016年第6期826-833,共8页Journal of Hydraulic Engineering

基　　金：国家自然科学基金重大计划重点支持项目(91547201);国家自然科学基金重大国际合作项目(51210014);国家重点基础研究发展计划(973计划)项目(2013CB035906)

摘　　要：将蛛群优化方法(SSO)这一新型群体智能寻优方法引入梯级水库群优化调度领域,并提出精英集聚蛛群优化方法(ESSO)进一步改善SSO的性能表现。ESSO在标准SSO方法寻优机制基础上,从精英个体动态更新策略与邻域变异搜索机制等两方面予以改进,提升蜘蛛群体的多样性与优秀个体领导能力,均衡方法的全局开采能力与局部勘探能力。澜沧江流域工程实例表明,所提ESSO方法能有效克服标准SSO的"早熟"缺陷,有效提升方法的搜索能力,可望为大规模梯级水库群优化调度提供一种新的高效求解思路。The novel meta-heuristic social spider optimization（SSO） provides a new effective way to solvethe multi-reservoir optimal operation problem. However,SSO has some disadvantages like premature conver-gence and poor search ability. Thus,the elite-gather social spider optimization（ESSO） is proposed to alle-viate the existing defects of the SSO. ESSO takes advantage of the elite individual to guide the evolution ofthe population for the globally optimal solution. Moreover,the neighborhood search mechanism is introducedinto ESSO to improve the local search ability of the method. The proposed approach is applied to the opti-mal operation of Lancang River cascade reservoirs,and the results shows that ESSO can improve the con-vergence property and robustness of SSO. Thus,ESSO is a feasible method for the multi-reservoirs opera-tion problem.

关 键 词：梯级水库群 优化调度 蛛群优化方法 精英导向 邻域搜索 变异机制 

分 类 号：TV697.1[水利工程—水利水电工程]