KLU稀疏直接求解器在状态估计中的应用

稀疏矩阵及稀疏线性方程组求解已成为大规模电网状态估计计算效率的瓶颈，阐释了基于BTF和Gilbert-Peierls算法的稀疏矩阵直接求解器KLU（Clark Kent LU），并将其嵌入到智能调度技术支持系统中生产运行的状态估计程序功能。首先在计算得到雅可比矩阵的基础上基于OpenMP并行化技术快速求解信息矩阵；然后使用KLU求解器进行信息矩阵的因子表符号分析和数值分解；最后在状态估计计算过程中使用KLU求解器求解线性方程组，从而提高大规模电网状态估计的计算效率。通过省中心、分中心、模型数据中心D5000的状态估计实际应用，证明了该方法的有效性和实用性。

Application of KLU Sparse Direct Linear Solver in State Estimation

Solving sparse matrix and linear equation is an important computational kernel in large-scale power system state estimation. This paper embeds the BTF and Gilbert-Peierls algorithm-based KLU sparse direct linear solver into the state estimation program of smart grid operation supporting systems. Firstly, based on the computation of the measurement jacobi matrix, the OpenMP parallelization technology is used to solve G which called gain matrix; and then, symbolic analysis and numerical factorization of G are handled by KLU solver; finally, the KLU solver is used to solve the linear equations in the process of state estimation, thus significantly improving the computational efficiency of state estimation in large-scale power network. The effectiveness and practicality of the proposed method are verified through practical application in D5000 system of provincial power company and model data dispatch centers.