基于改进遗传算法的溪洛渡水库优化调度研究  

Application of Improved Genetic Algorithm in Optimal Dispatching of Xiluodu Reservoir

作　　者：涂启玉[1] 梅亚东[1] 

机构地区：[1]武汉大学水资源与水电工程科学国家重点实验室,湖北武汉430072

出　　处：《水电能源科学》2008年第3期39-42,共4页Water Resources and Power

基　　金：国家重点基础研究发展计划(973项目)基金资助项目(2003CB415200)

摘　　要：针对溪洛渡水库年调度问题,建立了以发电量最大为目标的优化调度模型,分别利用GSA和NGA进行求解,所得结果与DP、GA、SA算法比较,结果表明改进后的GSA融合了GA全局搜索能力强和SA局部搜索能力强的特点,而NGA将GA个体置于特定的环境中优胜劣汰,增强了个体的多样性,两种改进遗传算法的收敛速度和计算结果都明显优于传统优化方法和基本遗传算法。Based on the analysis of the characteristic situation of Xiluodu Reservoir,a comprehensive optimal operation model is developed with consideration of its maximizing power generation of cascaded hydropower plants. This paper introduces Genetic Simulated Annealing and Niche Genetic Algorithms to solve the model of optimal operation. Compared to the dynamic programming ,basic genetic algorithm and Simulated Annealing algorithm,it is showed that GSA with the character of GA＇s stronger ability of overall search and SA＇s strong ability of local search, and NGA＇s placing the individual into particular environment to increase the variety of individual, so the Improved Genetic Algorithms have faster convergence velocity and better results.

关 键 词：遗传算法 遗传模拟退火算法 小生境技术 优化调度 

分 类 号：TV697.11[水利工程—水利水电工程]