基于多光源图像的接触网外绝缘健康状态评估

Health Condition Assessment of Catenary Insulation Based on Multisource Images

为了检测接触网外绝缘设备故障状态,采用紫外成像技术和红外热像技术对接触网外绝缘故障特征进行了研究。通过对接触网电气绝缘设备的现场检测,归纳出3种接触网典型的放电和发热故障现象。借助绝缘子局部放电试验,获取了不同污秽程度下的绝缘子的视在放电量和紫外图像;通过紫外图像特征提取,准确计算出放电光斑面积;根据视在放电量与紫外图像放电光斑的对比,得出了紫外光斑面积与视在放电量的正比关系。探究了紫外图像的拍摄距离、拍摄增益和拍摄仰角对放电光斑面积的影响,对不同拍摄参数下的光斑面积进行归一化处理并计算出其视在放电量的大小。通过试验探究了不同故障程度下的长时间温升规律:随着故障程度的增加,故障设备放电后最大温升越高,这为红外图像检测接触网外绝缘故障提供了依据。将紫外图像所检测到的视在放电量和红外图像所检测到的最大温升进行了模糊逻辑推理,据此构建了基于多光源图像的接触网外绝缘健康状态评估系统,经测试系统准确率为92.5%。因此可以认为,所构建的接触网外绝缘健康状态评估系统是合理有效的。

In order to detect the fault condition of the electric insulation equipment on catenary, we studied the fault characteristics of the electric insulation equipment by using ultraviolet imaging technology and infrared thermal imaging technology. Through on-site detection of the faulty electric equipment, we identified three typical discharge and heating failures on catenary. With the help of partial discharge tests of insulators, apparent discharge capacities and ultraviolet images of the faulty equipment under various filthy states were obtained. By extracting the characteristics of the ultraviolet images, the facula areas were accurately calculated. By comparing the capacities of partial discharge with the areas of facula on ultraviolet images, the direct proportional linear correlation between the two features was obtained. Moreover, we analyzed the impacts of shooting distance, gains and angles on the areas of facula, and used normalization to the areas of the facula under different parameters to figure out the apparent discharge capacities. By conducting tests, the regular pattern of temperature rise through a longtime under different fault conditions was generalized： the maximum temperature rise after discharging increased with the increment of fault degrees, providing a basis for detecting faults by using thermal imaging technology. The fuzzy logical inference should be conducted based on the discharge capacities detected by the ultraviolet images and the maximum temperature rise detected by the thermal images, and then an assessment system of the condition of catenary insulation based on ultraviolet and infrared imaging technologies has been established, and the accuracy rate is 92.5% in the test. Therefore, it can be confirmed that the catenary insulation health condition assessment system is effective.