大规模水电系统优化调度降维方法研究Ⅱ:方法实例  

Study on dimension reduction for optimal operation of large-scale hydropower system Ⅱ. Example for the method

作　　者：冯仲恺[1] 牛文静[1] 程春田[1] 武新宇[1] 申建建[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116023

出　　处：《水利学报》2017年第3期270-278,共9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：为缓解大规模复杂水电系统优化调度存在的"维数灾"问题,提出一种基于正交试验设计、在组合维实现降维的正交降维搜索方法。该方法将水电优化问题求解视为在发起点邻域内多次开展正交试验,选取"均衡分布、整齐可比"部分状态组合进行计算。分析表明该方法具有良好并行性,且时间复杂度和空间复杂度均为平方增长。同时,在应用至水电调度问题时,结合动态规划改进策略思想,提出4种不同的优化方法。贵州电网应用实例表明:所提方法具有计算参数少、鲁棒性强等优点,可有效均衡计算效率与求解精度,为大规模复杂水电系统优化调度的降维求解提供了新型方法。In order to solve the curse of dimensionality of joint optimal operation of large-scale hydropower system,orthogonal dimension reduction search algorithm（ODRSA） is presented on the basis of orthogonal experiment design.ODRSA solves the joint optimization scheduling problem of hydropower system by carrying out the orthogonal experimental design in the neighborhood of initial solution.During the iteration,some representative solutions are selected to calculate and the starting point will be replaced if the better solution is found.Analysis shows that the time complexity and space complexity of ODRSA are square.Simulation results of the cascade hydropower stations on the Wujiang River indicate that：ODRSA is a novel method for large-scale hydropower system with less calculation parameters,fast convergence speed and strong robustness.

关 键 词：水电 优化调度 正交试验设计 降维 维数灾 邻域搜索 计算复杂度 

分 类 号：TV697[水利工程—水利水电工程]