结合广度搜索的遗传算法在水库调度中的应用  

Genetic Algorithm Based on Deep Search and Its Application to Reservoir Optimal Operation

作　　者：张忠波[1] 张双虎[2] 蒋云钟[2] 

机构地区：[1]天津大学建筑工程学院,天津300072 [2]中国水利水电科学研究院,北京100038

出　　处：《南水北调与水利科技》2011年第5期85-88,共4页South-to-North Water Transfers and Water Science & Technology

摘　　要：为了改善遗传算法在水库优化调度中的应用效果,采用自适应遗传算法和广度搜索算子结合的算法,同时为保证水库优化调度搜索全局最优提供了一定保障。针对遗传算法容易陷入局部最优的缺点,引入正弦函数取随机数的广度搜索与遗传算法相结合的算法。通过分析比较单独使用自适应遗传算法或者广度搜索算法以及结合算法在实际水库优化调度中效果,结果显示,优化结果要比自适应遗传算法以及广度算法的结果更理想。充分证明了结合算法的高效全局搜索能力,避免了自适应遗传算法陷入局部最优,同时在一定程度上克服了广度搜索很难收敛的缺点,在一定收敛条件下得到了更接近全局最优的结果。To improve the solution searching efficiency of genetic algorithm in the optimization of reservoir operation, a deep searching operator combined with adaptive genetic algorithm was introduced to ensure locating global optima. To avoid the defect of genetic algorithm that might easily sink into local optima, an algorithm of deep searching operator was proposed with a sinusoidal random sampling function. The results of application in reservoir operation using the adaptive genetic algorithm combined with deep research operator and the new algorithm, respectively, were analyzed and compared, which showed that the new algorithm yielded a better result. The new algorithm had an efficient global searching ability, effectively avoided sinking into local optima, and further, overcame the shortcoming of slow convergence of the deep searching operator. Finally, the research obtained a better global optimal solution under a certain convergence criterion.

关 键 词：水库调度 遗传算法 广度搜索 正弦函数 自适应 变量罚函数 局部最优 收敛性 

分 类 号：TV697[水利工程—水利水电工程]