基于CAFSA的水库发电优化调度双层并行计算  

The Double-layer Parallel Calculation of Reservoir Generation Optimal Dispatching Based on Chaos Artificial Fish-Swarm Algorithm

作　　者：纪昌明[1] 吴昊[1] 黄锋[2] 张验科[1] 

机构地区：[1]华北电力大学可再生能源学院,北京102206 [2]中水珠江规划勘测设计有限公司,广州510610

出　　处：《中国农村水利水电》2016年第5期183-188,192,共7页China Rural Water and Hydropower

基　　金：国家自然科学基金资助项目(51279062;51179069;41340022);中央高校基本科研业务费专项资金资助项目(2014ZD12;13XS22;13XS23;13XS24;2010MS51)

摘　　要：应用于水库发电优化调度的人工鱼群算法(AFSA)后期搜索能力较弱,易陷入局部求解困境,并且运行时间相对较长;增加单体鱼的混沌优化搜索功能后,虽可提高最优求解精度,但计算耗时有所增加的问题也随之而出。依托互联多核计算机群,对水库发电优化调度中的混沌人工鱼群算法(CAFSA)展开双层并行计算架构设计,水库实例仿真的计算结果表明该并行策略能够有效保持CAFSA较高质量寻优的同时,可以大幅减少计算耗时,为提高调度模型求解效率提供了一定方向。The AFSA,applied in reservoir generation optimal dispatching,is weak in searching capacity at the later searching stages.It is easy to meet with the difficulty in getting partial solution and it runs for a relatively long time.The precision of optimal solution can be improved by increasing the chaotic optimal searching function of one single fish,but the problem of computing time increment is produced.Based on multi-core computers interconnected,a double-layer parallel computing architecture is designed for CAFSA in reservoir generation optimal dispatching.The computing results of reservoir instance simulation show that the parallel strategy can effectively maintain the high quality of CAFSA.It can greatly reduce the computing time and provide a certain direction for improving the solving efficiency of the scheduling model.

关 键 词：双层并行计算 混沌优化 人工鱼群算法 水库 发电优化调度 

分 类 号：TV697.1[水利工程—水利水电工程] TP301.6[自动化与计算机技术—计算机系统结构]