基于FCM算法的连锁故障评价及路径划分方法研究

Cascading Failure Evaluation and Path Division Based on FCM Algorithm

提出了一套新的连锁故障评价指标,以此来反映不同的随机因素组合引发连锁故障的可能性及蔓延速度,并研究了如何根据分立指标对已知样本数据进行综合评价和划分以实现样本数据筛选。利用多目标的分析方法,首先对样本数据进行综合评价,然后以样本数据的分立指标和综合指标评价作为特征输入采用FCM算法进行划分。最后在IEEE39系统上进行算例分析,验证了评价方法的合理性。

Cascading failure is one of the main causes of blackouts. A new set of cascading failure evaluation is proposed to reflect the possibility and spreading speed of cascading failures, which are caused by different random combinations of factors. The issue that how to evaluate and classify data samples to achieve the sample data screening by separate indexes is studied. By using a multibjective analysis method, the data samples are evaluated, and the FCM algorithm is adopted to classify sample data with separate and comprehensive indexes as input. The evaluation method is proved to be reasonable by an example in IEEE39 system.