基于S变换和时域分析的电能质量扰动识别

Power Quality Disturbances Classification Based on S-Transform and Time Domain Analysis

提出了一种基于S变换(S-Transform,ST)和时域分析的电能质量扰动自动识别方法,该方法利用信号的S变换幅值矩阵和时域信息快速提取与各类电能质量扰动相应的特征,并借助简单的规则树识别特定的扰动,避免了因训练样本不足引起的较大误差,提高了识别效率。仿真试验结果表明,该方法识别率高,抗噪能力强,适用于电能质量扰动监测和辨识系统。

The authors propose a new approach to detect and classify power quality disturbances by combining the S-transform with time-domain analysis, in which the time-domain information and amplitude matrix of S-transform of signals are used to rapidly extract the features corresponding to various disturbances in power quality, and by means of simple rule-based classification tree the specified disturbance can be recognized to avoid the fair-sized error caused by insufficient training samples, therefore the classification efficiency can be improved. Simulation results show that the proposed method possesses high recognition rate and strong resistances to noises, so it is suitable to the monitoring and classification system for power quality disturbance.