基于红外图像的低压配电柜内部短路故障识别研究  

Research on Internal Short Circuit Fault Identification of Low-Voltage Distribution Cabinet Based on Infrared Images

作　　者：曹家君 李秋雨 周儒畅 CAO Jiajun;LI Qiuyu;ZHOU Ruchang(State Grid Shanghai Fengxian Electric Power Supply Company,Shanghai 201499,China)

机构地区：[1]国网上海奉贤供电公司,上海201499

出　　处：《自动化应用》2024年第14期171-174,共4页Automation Application

摘　　要：由于低压配电柜内部结构和电气元件复杂,短路故障的信号特征可能受多种因素的干扰,导致识别偏差值较大。为此,进行基于红外图像的低压配电柜内部短路故障识别研究。首先,预处理低压配电柜内部的红外图像,以提高图像质量和对比度。其次,利用预处理后的红外图像,通过图像分割和特征提取技术,定位短路故障发生的区间,缩小故障搜索范围。在此基础上,构造节点故障特征数据,提取关键故障特征量,以量化描述故障的特征。最后,通过对比正常与故障状态下的特征量差异,实现低压配电柜内部短路故障的准确识别。结果表明,在区间AB段的任何地方,与基于电流分析的方法和基于电气信号波形分析的方法相比,基于红外图像的低压配电柜内部短路故障识别方法的偏差值都是最小的,故障识别结果和实际故障所处的区间一致,实际应用效果更好。Due to the complex internal structure and electrical components of low-voltage distribution cabinets,the signal characteristics of short-circuit faults may be affected by various factors,resulting in significant identification deviation values.For this purpose,infrared image-based short circuit fault identification inside low-voltage distribution cabinets is investigated.Firstly,preprocess the infrared images inside the low-voltage distribution cabinet,to improve image quality and contrast.Secondly,using preprocessed infrared images and image segmentation and feature extraction techniques,locate the interval where short circuit faults occur and narrow down the fault search range.On this basis,construct node fault feature data and extract key fault feature quantities to quantitatively describe the characteristics of faults.Finally,by comparing the difference in characteristic quantities between normal and fault states,accurate identification of internal short circuit faults in low-voltage distribution cabinets can be achieved.The results show that at any point in the AB section of the interval,compared with the methods based on current analysis and electrical signal waveform analysis,the deviation value of the internal short circuit fault identification method based on infrared images in low-voltage distribution cabinets is the smallest.The fault identification results are consistent with the interval where the actual fault is located,and the practical application effect is better.

关 键 词：红外图像 低压配电柜 短路故障识别 内部短路 

分 类 号：TM472[电气工程—电器]