摘要
鉴于巨型水电站群短期联合调度计算的高精度和高时效要求,在多核PC机群环境下,设计了基于水电站群拓扑结构与优化算法寻优轨迹的双层并行优化方法,并采用MPI+OpenMP混合细粒度模型,实现了多进程与多线程的同步并行计算,并结合不同电站数和PC机群规模,对并行优化算法的性能进行了测试和对比分析。实例应用结果表明,该算法在确保精度的前提下可大幅缩减耗时,显著提高计算效率。
For the high accuracy and timeliness requirements of short-term joint dispatch of giant-scale hydropower station groups, double parallel optimization method based on topological structure of hydropower station groups and opti- mization trajectory of algorithm is designed with multi-core PC clusters. Multi process and multi-thread synchronous par- allel computing are implemented with mixed fine-grained model of MPI and OpenMP. Finally, this article analyzes the calculation performance of algorithm with different numbers of hydropower station and process. The instance results show that the calculation time is reduced greatly on the premise of ensuring optimization accuracy, and the computing ef- ficiency is improved significantly.
出处
《水电能源科学》
北大核心
2013年第11期45-48,167,共5页
Water Resources and Power
基金
国家电网公司科技基金资助项目(SG10011)
关键词
水电站群
短期联合调度
并行优化
多核PC机群
hydropower station groups
short-term joint dispatch
parallel optimization
multi-core PC clusters
