基于定性趋势分析的无绝缘轨道电路电气绝缘节设备故障诊断方法

Equipment Fault Diagnosis for Electrical Separation Joint of Jointless Track Circuits Using Qualitative Trend Analysis

根据均匀传输线理论建立无绝缘轨道电路的机车信号感应电压幅值包络计算模型,分析电气绝缘节设备故障、补偿电容故障、道砟电阻及列车运行速度等因素对机车信号感应电压幅值包络曲线的影响规律。以临近发送端3个补偿电容间距的机车信号感应电压幅值包络曲线为分析对象,提出基于定性趋势分析理论的电气绝缘节设备故障诊断方法。首先以区间半分法提取信号趋势信息;然后通过趋势匹配算法计算实际信号趋势与诊断规则知识库中各故障模式信号趋势的相似度;最后通过相似度值间的比较进行故障诊断。仿真实验结果表明:该方法对电气绝缘节设备的早期故障有很好的诊断效果,当信噪比为50dB时,故障诊断正确率可达94.4%,且对道砟电阻的波动具有很强的适应性。

The induced voltage amplitude envelope of cab signal was simulated according to the calculation model of a jointless track circuit based on uniform transmission line theory.The influence laws of such factors as the equipment faults of electrical separation joints,the faults of trimming capacitors,the variations of ballast resistance and train running speed,etc.on the induced voltage amplitude envelope curve of cab signal were investigated.With the induced voltage amplitude envelope curve of cab signal between the transmitter and the third trimming capacitor as the analysis object,a fault diagnosis approach for the electrical separation joint based on qualitative trend analysis theory was proposed.Signal trend information was extracted using interval-halving algorithm,and then trend matching algorithm was adopted to calculate the similarity between an actual signal trend and that under each fault mode in the knowledge base of diagnosis rules.Finally,fault diagnosis was done through comparing similarity values.Simulation experiment results show that the incipient faults of the electrical separation joint are well detected using this approach.The correct rate of fault diagnosis can reach 94.4％ when SNR is 50 dB,and it has higher adaptability to the variations of ballast resistance.