Abstract: In order to monitor the pollutant emission indicators of heavy-duty diesel vehicles in real time and quickly screen vehicles with excessive emissions, the Apache open-source big data framework with Flink as the core computing engine is used to achieve real-time processing of the data sent back by the heavy-duty vehicle emission on-board terminal and calculate the vehicle pollutant emission level. Based on the calculation results, the information of vehicles with excessive emissions will be pushed to vehicle users, after-sales departments, and research and development departments. At the same time, according to the requirements of the demand department, the operating conditions, configuration information, fault code information of the vehicles exceeding the standard are collected and analyzed to quickly locate the causes of excessive emissions. This will effectively improve social and economic benefits.

Key words: Stream processing; Emission warning; Heavy duty diesel vehicles; Flink

摘要： 为了能够实时监控重型柴油车污染物排放水平，快速筛查出排放超标的车辆。利用以 Flink 计算引擎为核心的 Apache 开源大数据框架，根据重型车排放车载终端发回的数据计算 出车辆污染物排放水平。开发完成后，该预警监控系统会根据计算结果向车辆用户、售后部 门、研发部门推送排放超标车辆信息，同时根据需求部门的要求，统计超标车辆的运行工况、 配置信息、故障码信息等，实现快速定位排放超标原因，有效提高了社会和经济效益。

关键词: 流处理；排放预警；重型柴油车；Flink