基于物联感知的城轨设备智能采集与监控优化OACSTPCD
INTELLIGENT COLLECTION AND MONITORING OPTIMIZATION OF URBAN RAIL TRANSIT EQUIPMENT BASED ON INTERNET OF THINGS PERCEPTION
针对泛在物联网中城市轨道交通大规模机电设备的信息采集和展示低效以及性能瓶颈问题,提出一种感知设备的大数据采集和智能化处理方法.通过数据点自辨识、自适应、递阶连接,实现设备状态捕获的批处理.设计图元图像识别智能算法自动生成监控画面.对实时采集数据的逻辑运算和语义模型转换,拟合多组数据综合反映关联系统状态,定义分级分类报警策略,实现数据驱动的智能监控.经验证,接口标准化大幅加速批量设备的接入,优化算法改进图纸识别率达90%,连接对象并准确显示的可用性达85%以上,明显提高监控效率和正确率.
Aimed at the inefficiency and performance bottleneck of information collection and display of large-scale electromechanical equipment in urban rail transit(URT)ubiquitous in the Internet of things,a method of massive data collection and intelligent processing of perceptual devices is proposed.Through self-identification,self-adaptive and hierarchical connection of data points,batch processing of the equipment status capture was realized,and the intelligent algorithm of graphic and image recognition was used to automatically produce the equipment monitoring interface.According to the logical calculation of real-time data and semantic model transformation,the combined values of multiple groups of data were fitted to reflect the comprehensive state of associated systems and further define the monitoring strategy of classified alarm,to realize data driven intelligent monitoring.The experimental results show that data interface standardization can significantly speed up the access a great quantity of equipment,and the optimized algorithm can make the recognition rate of system drawing to 90%,also,the availability of connecting objects and display reaches more than 85%.The efficiency and accuracy of the monitoring are improved.
张铭;王石生;高凡
中国铁道科学研究院集团有限公司电子计算技术研究所 北京 100081
计算机与自动化
城市轨道交通物联网监测感知图像识别批量化
Urban railtransitInternet of ThingsMonitorPerceptionImage recognitionMass production
《计算机应用与软件》 2024 (001)
26-35 / 10
科技部国家重点研发计划项目(2018YFB1201404);北京市自然科学基金-丰台轨道交通前沿研究联合基金项目(L221010);中国铁道科学研究院集团有限公司科技研究开发计划项目(2020YJ192).
评论