Probabilistic Load Flow Method Based on Modified Latin Hypercube-Important Sampling

Abstract

The growing amount of distributed generation has brought great uncertainty to power grids. Traditional probabilistic load flow (PLF) algorithms, such as the Monte-Carlo method (MCM), can no longer meet the needs of efficiency and accuracy in large-scale power grids. Latin Hypercube Sampling (LHS) develops a sampling efficiency and solves the correlation problem of distributed generation (DG) access nodes for accuracy analyses. In this paper, a modified Latin Hypercube-Important Sampling method is proposed for higher efficiency and precision by using the importance sampling method before LHS and the Cholesky decomposition method in correlation calculations. The simulation results are presented using a modified IEEE 30-bus system and are compared with traditional MCM and LHS.