基于多源信息融合的设备动作状态自动识别方法的研究

摘要

The paper introduces an automatic identification method of substation high voltage breakers and other equipment action state based on multi-source information fusion. Firstly, the paper introduces the cur-rent automatic identification methods and expounds advantages and disadvantages of each method;besides, it proposes an automatic action state identification method of substation high voltage breakers and other equip-ment based on multi-source information fusion, namely a fusion technology of monitoring system based remote identification, infrared identification of inspection robot as well as video and monitoring linked identification to quantize the identification result of high voltage breakers. Through quantification index matching technique, the linked high voltage breaker action state identification by the three methods is implemented to enhance i-dentification accuracy. The experiment result shows that the multi-source information fusion based method can effectively and accurately identify the action state of substation high voltage breakers and other equipment.%提出了一种基于多源信息融合技术对变电站高压断路器等设备的动作状态进行多层次自动识别的方法。首先介绍当前对高压断路器等设备的动作状态进行识别的方法，分别指出各方法的优缺点，在此基础上提出了针对高压断路器等设备进行多源信息融合技术的自动识别方法，即采用监控系统远程识别、巡检机器人的红外识别和视频监控联动识别的融合技术，量化高压断路器等设备的识别结果，并采用量化指标的匹配技术，实现3种方法对高压断路器等设备动作状态的联动识别，增强识别结果的准确性。实验结果表明，基于多源信息融合的变电站高压断路器等设备动作状态自动识别方法可有效、准确地识别变电站内设备的动作状态。