基于MCMC法进行电压跌落随机预估方法的研究

Research on stochastic estimation of voltage sag based on MCMC method

针对电压跌落随机预估中蒙特卡洛法(MC法)静态性、计算效率低、耗时长的缺陷,提出基于马尔可夫链蒙特卡罗法(Markovchain Monte Carlo,MCMC)对电压跌落进行随机预估。研究建立了电压跌落故障状态变量的数学模型,并利用Matlab建立了IEEE-9节点测试系统模型,使用Gibbs抽样法得到故障模型的状态变量。通过分析电压跌落幅值的概率分布,并用MCMC法和MC法分别对电压跌落指标进行仿真计算。仿真结果表明,该方法较蒙特卡罗法稳定性好,收敛速度快,计算时间短。

The Monte Carlo （MC） method in the stochastic assessment of voltage sags suffers from low computing efficiency, static characteristic and long time consuming. According to those defects, the paper presents a stochastic assessment based on Markov chain Monte Carlo （MCMC） method. A mathematical model of state variables of voltage sags fault is built up, an IEEE nine nodes testing system model is put up in Matlab, and state variables of the fault model are obtained by Gibbs sampling method. The paper analyzes the probability distribution of the amplitude of voltage sags, simulates the indicator of voltage sags by MCMC and MC method respectively, and verifies the feasibility of the MCMC method. The simulation results show that, this method has better stability, faster convergence rate and shorter calculation time compared with MC method.