电力系统潮流并行计算中的方程组求解方法

摘要

随着高性价比并行计算系统的技术发展,大规模电力系统潮流并行计算应用越来越广泛.针对电力系统潮流并行计算中方程组求解这一核心问题,首先分析了潮流并行计算大型矩阵常用算法,通过对稀疏矩阵压缩算法以及矩阵分块算法进行性能比较,给出了不同算法的适用范围.其次,探讨了潮流计算线性方程组求解并行算法,对比分析了共轭梯度法、GMRES等不同算法的优点和局限性.并对进一步提高电力系统潮流并行计算效率的研究方向进行了展望.%With the development of cost-effective parallel computing system technology,parallel computing of power flow in large scale power system is being widely used.This paper focuses on the core technology of the equations solving in the parallel com-puting of power flow.Firstly,the common algorithms of large matrices in parallel computing of power flow is analyzed,and sparse matrix compression algorithm and matrix partition algorithm are compared,thus the scope of the application of different algorithms are given.Then the parallel algorithms of power flow about solving the linear equations are discussed.The advantages and limitations of different algorithms such as conjugate gradient method and GMRES are also compared and analyzed.Finally,this paper looks at the prospects for future research directions of the parallel computing efficiency of power system.