基于无人机多传感器数据采集的电力线路安全巡检及智能诊断

Safety Inspection and Intelligent Diagnosis of Transmission Line Based on Unmanned Helicopter of Multi Sensor Data Acquisition

为满足电力线路应急处置和安全巡检业务的高效、自动化处理需求,设计研发了一套无人直升机多传感器电力线路安全巡检系统。首先,通过高精度同步授时与标定,统一机载紫外、红外、可见光以及激光传感器的时间、空间基准。其次,无人机通过机载多传感器同步采集输电线路的高精度3维激光点云、高分辨率航空影像、红外视频、紫外视频,进行多源数据的独立与融合处理,完成对线路通道的安全距离检测、线路设备异常发热及异常放电检测,实现对输电线路不同缺陷和隐患的智能诊断。最后输出标准诊断报表,以供现场勘察确认。为验证无人机多传感器巡检系统的功能,开展了实际带电运行线路巡检试验。巡检结果表明,无人机巡检系统对线路缺陷与隐患的检测结果与人工现场勘察结果一致。该系统能发现输电线路多种运行缺陷和安全隐患,为运行单位提供相应的检修维护决策依据,还能提供缺陷或隐患位置的可视化场景信息,方便人工辅助诊断和排查,提高输电线路巡检效率。

To meet the demand.of effective, automated power-line emergence management and maintenance inspection, we developed a multi-sensor power-line inspection system of unmanned helicopter. Firstly, through precise time synchronization and calibration, the system unifies the time and space criteria of its sensor, including ultraviolet sensors, infrared sensors, visible-light sensors, and laser sensors. Then, the multiple on-board sensors collect images of transmission lines simultaneously including precision 3-D laser point clouds, high-resolution aerial images, infrared video images, and ul- traviolet video images. These fused multi-source data are put through intelligent diagnosis reciprocally and independently to accomplish the detection of defect and hazard on power lines, as well as the detections of safe distance and abnormal heat. Finally, diagnosis reports are given for on-the-spot investigation. We conducted inspection tests of actual lines to ve- rify the proposed unmanned helicopter inspection system＇s functions. It is obtained that the system achieves the same diagnosis results as that made by skilled power-line inspection technicians. The system is able to find multiple kinds of defects and safety risks of transmission lines and accordingly provide decision basis for operation companies. It also of- fers visualization models and scene information of defects or hazard positions, which is useful for artificial auxiliary diagnosis and improving the efficiency and reliability of fault detections.