基于交叉熵重要抽样和状态筛选的电力系统可靠性计算方法

Reliability Calculation Method of Power System Based on Cross Entropy Importance Sampling and State Screening

针对蒙特卡洛法模拟在计算电力系统可靠性时效率比较低的问题,提出一种基于交叉熵重要抽样和状态筛选的电力系统可靠性计算方法。首先,利用交叉熵重要抽样计算电力系统的修正概率分布及各元件的修正强迫停运率。其次,确定各元件的重要程度,根据重要元件确定对系统可靠性指标影响较大的部分系统状态,从而形成重要状态子空间。最后,对剩余的系统状态采用修正概率分布进行抽样,在抽样过程中对重要状态和已被抽样的系统状态进行筛选,避免已被分析过的系统状态被重复多次分析。该方法既利用交叉熵重要抽样提高了抽样效率,又避免了对同一状态的重复分析,在状态抽样和状态分析2个步骤上的计算效率均有提升,IEEE-RTS系统验证了所提方法的效率。

In terms of the low efficiency of Monte Carlo simulation in calculation of power system reliability,a method of power system reliability calculation cross entropy importance sampling and state screen was proposed.Firstly,the modified forced outage rates and probability distribution function were calculated with CEIS.Then,the importance sub-space formed by important states which determined by important components,the left states would be sampled by modified PDF.The important states and sampled states would be screened out to avoid repetitive calculation for the same state.The method improves sampling efficiency as well as avoides redundant calculation of the same state.The increasing the efficiency of state sampling and state calculation,which was proved in IEEE-RTS system.