基于遗传-蚁群算法的库群长期优化调度  

Genetic-Ant Algorithm for Long-term Optimal Operation of Hydropower Systems

作　　者：汪明清[1] 牛文静[2] 廖胜利[2] 

机构地区：[1]贵州电网公司电力调度控制中心,贵州贵阳550002 [2]大连理工大学建设工程学部,辽宁大连116024

出　　处：《水电能源科学》2013年第11期49-52,194,共5页Water Resources and Power

基　　金：国家自然科学基金资助项目(51209029);高等学校博士学科点专项科研基金资助项目(20120041120002)

摘　　要：为克服蚁群算法前期收敛慢、易陷入局部最优解且参数难以确定的缺点,提出了遗传-蚁群算法进行库群长期优化调度问题的求解。该方法应用遗传算法生成问题的初始解,并将初始解的适应度转化为蚁群算法的信息素初始值,同时引入遗传算法染色体交叉、变异的思想进行蚁群算法参数最优组合的确定,提高了蚁群算法的优化性能和求解精度。对乌江流域4座水电站的计算结果表明,该算法可显著改善优化结果质量,获得良好的调度方案,是求解库群长期优化调度问题的一种有效方法。In order to overcome the weakness of early slow convergence, easily trapping into local optimal solution and with difficulty to determine parameters existing in ant colony algorithm, Genetic Ant method is proposed to solve long-term optimal operation of large-scale hydropower system. This algorithm takes advantage of GA to generate initial solution and transform its fitness into original value o{ pheromone in ant colony. Meanwhile, the idea of crossover and mutation of Chromosome in GA is introduced to achieve parameters optimization combination of ant colony, which im- proves the performance and accuracy of the algorithm. The results of 4 hydropower stations in Wuiiang River show that Genetic Ant method can improve the quality of the optimization solution and contribute to better scheduling. Therefore, this algorithm is effective in solving long-term optimal operation of hydropower systems.

关 键 词：水电站群 优化调度 遗传一蚁群算法 最优路径 

分 类 号：TV213.4[水利工程—水文学及水资源]