基于双阶段并行隐马尔科夫模型的电力系统暂态稳定评估

Power System Transient Stability Assessment Based on Two-stage Parallel Hidden Markov Model

基于人工智能机器学习的暂态稳定评估越来越成为研究热点,提出一种基于双阶段并行隐马尔科夫模型(two-stage parallel hidden Markov model,TS-PHMM)的电力系统暂态稳定评估精细化模式识别方法。第1阶段采用相对灵敏度对原始电气特征量进行筛选,找出对电网动态变化敏感度高的特征子集;第2阶段采用主成分分析对特征子集进行排序,得到能够反映电网动态响应特性且线性无关的最优特征子集;最后,通过并行隐马尔科夫模型训练对暂态稳定进行模式识别。在CEPRI 8机36节点以及实际区域电网环境上的仿真分析,验证了该方法的有效性和精确性。在辨识准确率相当的情况下,该方法比常用人工智能类方法(如ANN,SVM等)所需训练样本更少、收敛更快。

The transient stability assessment （TSA） of power system based on the artificial intelligence and machine learning method has become more popular. This paper proposed a precise pattern recognition method for TSA based on a two-stage parallel hidden Markov model （TS-PHMM）. In the first stage, the sensitive feature subset was selected from the original feature set based on the relative sensitivity principle; in the second stage, the principal component analysis （PCA） method was used to decrease the subset dimension to obtain an optimized feature set. Then the optimized subset was adopted to train PHMM with a serial weight factors for TSA. Finally, in the CEPRI 8-generator 36-bus test system and a real large power system, the simulation results proved the validity and effectiveness of the feature selection approach and PHMM pattern recognition. Meanwhile, this new method needs less training samples compared with some of the common methods such as SVM and ANN to reach an equivalent accurate rate.