基于神经网络的重型车辆远程监控NOx传感器露点保护过程数据修复方法
Neural Network-Based Data Repair Method During NOx Sensor Dew Point Protection in Remote Monitoring of Heavy-Duty Vehicles
刘春涛 1张帆 2吴春玲 3裴毅强 2陈淑鑫 1何颖1
作者信息
- 1. 天津仁爱学院 机械工程学院,天津 301636
- 2. 天津大学 机械工程学院,天津 300072
- 3. 中汽研汽车检验中心(天津)有限公司,天津 300300
- 折叠
摘要
为解决重型车辆远程监控数据中NOx传感器露点保护过程的数据无效问题,利用一辆国六重型车辆的PEMS测试对露点保护期间的高NOx排放问题进行探究,验证了利用神经网络算法修复数据和提高远程监测数据利用率的可行性.结果表明,NOx传感器露点保护过程会导致30%以上的NOx排放量未被统计;在露点保护期间,超过90%的数据显示车辆速度低于54 km/h、发动机冷却液温度低于82℃、SCR入口温度低于245℃、SCR出口温度低于225℃.神经网络算法可有效修复露点保护过程中失效的NOx测量值,对发动机原始排放和车辆尾管排放的累计排放量误差都在4%以内.
Abstract
To solve the problem of invalid data during the dew point protection phase of NOx sensors in the remote monitoring of heavy-duty vehicles,the paper used the PEMS tests on a China VI heavy-duty vehicle to investigate the high NOx emissions during this protection period.Furthermore,the feasibility of using a neural network algorithm to repair the data and improve the utilization rate of remote monitoring data was verified.The results show that the dew point protection leads to more than 30%NOx emissions not being recorded.During this protection phase,over 90%of the data revealed that the vehicle speed was below 54 km/h,the engine coolant temperature was below 82 ℃,the SCR inlet temperature was below 245 ℃,and the SCR outlet temperature was below 225 ℃.The neural network algorithm effectively repaired the invalid NOx measurements during dew point protection,with errors of less than 4%.
关键词
神经网络/远程监控数据/NOx排放/重型车/露点保护Key words
neural network/remote monitoring data/NOx emissions/heavy-duty vehicles/dew point protection引用本文复制引用
基金项目
国家重点研发计划(2022YFC3701805)
国家重点研发计划(2022YFC3703600)
出版年
2024
国家科技期刊平台
NSTL
维普
万方数据