The existing annual inspection and emission test cycle of motor vehicles have not effectively curbed diesel vehicle emissions, as real-world emissions can far exceed regulatory limits due to issues such as aging of aftertreatment systems and manipulation of sensor data. The China VI emission standard for heavy-duty diesel vehicles requires the installation of a remote on-board unit to upload real-time driving data to an online monitoring platform for emission regulation. The development of corresponding data analysis and application models is urgently needed. In this paper, the portable emission measurement system was employed to conduct real drive emissions tests on 3 heavy-duty diesel vehicles to obtain test data. The maximal relevance and minimal redundancy algorithm was applied to extract input features, and then the NOx transient emission prediction model was established using categorical boosting. The root mean square error of the prediction model on the test set is 1.034, with an average absolute error of 0.3907, an average absolute percentage error of 0.1443, and a coefficient of determination of 0.9047. Shapley additive explanation was applied to explain the feature importance of the model. This article offers useful insights for the online supervision of NOx emissions from heavy-duty diesel vehicles.
Remote NOx emission monitoring for heavy duty diesel vehicles using categorical boosting
Third International Conference on Electronic Information Engineering, Big Data, and Computer Technology (EIBDCT 2024) ; 2024 ; Beijing, China
Proc. SPIE ; 13181
2024-07-19
Conference paper
Electronic Resource
English
On-Road Emission Factors for Heavy-Duty Diesel Vehicles
British Library Conference Proceedings | 1994
|Emission performance of paraffinic HVO diesel fuel in heavy duty vehicles
SAE Technical Papers | 2011
|Emission performance of paraffinic HVO diesel fuel in heavy duty vehicles
British Library Conference Proceedings | 2011
|