基于突变理论信息融合的故障电弧检测方法

Arc-fault Detection Method Based on Information Fusion of Catastrophe Theory

根据故障电弧产生时电流波形发生突变的特征,提出运用突变理论信息融合算法对低压开关柜(箱)内故障电弧检测的方案。通过积分、均方根运算,小波分析和频谱分析对电流时域、时频域和频域特征进行提取。采用积分值变化系数、相邻周期均方根值、高频变化系数和3次谐波分量变化量4个特征量表征电流突变,根据突变原理建立故障电弧判别模型,对特征量信息融合求出故障电弧判别指标。结果表明,电弧故障前后2个周期电流特征量及故障电弧判别指标相比正常运行时明显增大,通过设定阈值和多次检测能有效判别不同负载类型电弧故障及区分负载改变电流,有很好的泛化性和鲁棒性。

According to the mutation of current waveform when arc-fault occurred,a detection method for low-voltageswitchgear（box）arc fault based on information fusion of catastrophe theory emploed was proposed. Integration,rootmean square（RMS）formula,wavelet analysis and spectrum analysis are used to extract features of current in time do-main,time-requency domain and frequency domain. At this process,four characteristics which are variation coefficientof the integral value,RMS between two adjacent cycles currents,variation coefficient of high-frequency coefficients andthird harmonic component variation were defined to indicate the amount of mutation. They were taken for information fu-sion by establishing the evaluation model of arc fault based on mutation principle to obtain arc fault evaluation index.The results show that the characteristics and evaluation index of two cycles current before and after arc fault occurredare significantly increased compared to normal operation,which can effectively distinguish fault arc and differentiatethe current when the load changes with good generalization and robustness.