摘要
电客车日常巡检的效率与准确性是城市轨道交通安全运营的重要保障。针对传统人工巡检效率低、可靠性差等问题,本研究开发了基于多模态数据融合的城市轨道交通电客车智能巡检系统;研制了集成声音、图像、温度的多模态数据采集传感器;通过将多模态数据有效融合,进行了电客车车底项点状态数据的智能检查;围绕多模态数据智能配准、高光去除图像增强、自适应图像分割等算法开发,构建了软硬件结合的电客车智能巡检系统,并在杭州、宁波等多地的城市轨道交通电客车智能巡检中进行了应用,实现了螺栓、油液位等电客车底部27类检测项点的智能巡检,提高了城市轨道交通行业故障检测的时效性和准确性。
The efficiency and accuracy of daily inspection of electric buses is an important guarantee for the safe operation of urban rail transit.Aiming at the problems of low efficiency and poor reliability of traditional manual inspection,the study develops an intelligent inspection system for urban rail transit electric buses based on multimodal data fusion.A multimodal data acquisition sensor integrating sound,image and temperature is developed.Through the effective fusion of multimodal data,the intelligent inspection of the state data of the items on the bottom of the electric bus is carried out.Based on the intelligent registration of multimodal data,image enhancement of highlight removal,and the development of adaptive image segmentation algorithm,an intelligent inspection system for electric buses combining software and hardware is developed,which has been applied in the intelligent inspection of electric buses in urban rail transit in Hangzhou,Ningbo City and other cities.The intelligent inspection of 27 types of inspection items at the bottom of electric buses,such as bolts and oil level,is realized,and the accuracy and timeliness of anomaly and fault detection in urban rail transit industry are improved.
出处
《科技风》
2026年第9期10-12,共3页
基金
杭州市重点研发计划项目(2024SZD1A26)。
关键词
城市轨道交通
电客车
智能巡检
多模态传感器
urban rail transit
electric bus
intelligent inspection
multimodal sensor
作者简介
周思杭(1981-),男,汉族,浙江杭州人,博士研究生,高级工程师,研究方向为轨道交通装备;通信作者:彭翔(1989-),男,汉族,河南信阳人,博士研究生,副教授,研究方向为结构优化设计。
