基于多元信息的绝缘子红外故障诊断

绝缘子在制作不良、污秽等故障情况下表现为温度升高,红外成像技术是绝缘子温升故障检测的一种新手段。针对绝缘子的温升故障,提出了基于多元信息的图像识别和故障诊断方法。通过对图像预处理,利用边缘检测和数学形态学算法去除红外图像中的复杂背景,将绝缘子准确的识别出来,再结合红外图像的温度信息,判断绝缘子的温度是否超出设定的阈值,实现绝缘子的故障诊断。利用红外摄像仪获取故障绝缘子的红外图像,验证了本方法的可行性。

Infrared Fault Diagnosis of Insulator Based on Multi-information

Insulators would have temperature rising under fault conditions such as low-value and contamination.Infrared thermography is a new method to detect the temperature rising fault.An integrated method for the insulator fault diagnosis was proposed by using multi-information including both image and temperature features.After filtering the noise in infrared image,edge detection and morphology method were applied to eliminate the complicated background and extract the insulator from the infrared image.With the precise edge of insulator,its working condition could be judged by comparing its temperature with a threshold.Finally,more infrared images prove the feasibility of the proposed method.