基于弓网监测的受电弓缺陷智能检测技术研究

Study on the Intelligent Detection Technology for Pantograph Faults based on Pantograph-Contact Line Monitoring

城轨车辆在运行过程中,受电弓的工作状态是保证安全运营的重要因素。目前大多检测方法仍停留在人工检测阶段,存在效率低、无检测标准等问题。近年来随着人工智能技术的高速发展,解决了许多传统技术难以突破的难题。文章结合深度学习、机器学习和统计方法,提出了一种基于弓网视频分析的受电弓缺陷智能检测技术,该项技术可在嵌入式平台中部署智能识别算法模型,利用采集的视频监控图像进行实时处理、检测并识别受电弓故障缺陷。主要针对受电弓异物、弓角异常等常见故障进行了理论研究,并通过试验验证了该技术的实时性和有效性。

During the operation of the urban rail trains,the working state of pantographs is an important factor to ensure safe operation.At present,most detection methods still remain in the manual detection phase,with problems such as low efficiency and a lack of detection standards.With the high-speed development of AI technology in recent years,many difficult problems insurmountable by conventional technologies have been solved.The article combines deep learning,machine learning,and the statistical method to propose an intelligent detection technology for pantograph faults based on an analysis of pantograph-contact line videos.This technology can deploy an intelligent identification algorithm model on the embedded platform and use the collected video monitoring images for real-time processing,detection and identification of pantograph faults and defects.The article bases the theoretical study mainly on common faults,such as foreign bodies in pantographs and abnormalities of pantograph horns,and verifies the real-time performance and effectiveness of this technology by testing.