基于小样本机器学习的电力巡检技术研究

Research on Power Patrol Inspection Technology Based on Small Sample Machine Learning

本文设计出通过无人机进行电路巡检的技术方案,该方案通过GPS定位和路径规划使无人机在特定位置采集电力传输设备运行图像数据,并通过本文改进的小样本目标识别算法对采集到的图像信息进行识别,实现了潜在故障和安全隐患问题分析。试验结果表明,本研究方法分类的正确率提高了29.15%,样本分析提高了18.63%,提高了检验的成功率。

This paper designs a technical solution for circuit inspection by UAVs.The program uses GPS positioning and path planning to enable UAVs to collect operating image data of power transmission equipment at specific locations,and uses the improved small sample target recognition algorithm to collect data.The received image information is identified,and the analysis of potential faults and hidden safety hazards is realized.The experiment results show that the classification accuracy of the method has increased by 29.15%,sample analysis has increased by 18.63%,and the success rate of inspection has been improved.