Research On Real-world No_x Emission Of A China Ⅵ(b) Heavy-duty Diesel Vehicle

Heavy diesel vehicles were the main contributors to NO_x emissions from mobile pollution sources,China Ⅵ standard put forward higher requirements for NO_x emission limits and the durability of related devices.At the same time,it had intensified the supervision of in-use compliance and for the first time the emissions warranty period was set.At present,the deterioration of heavy diesel vehicles’NO_x emission was studied mainly based on the rapid aging bench test or the on-road vehicle durability test.Both methods need to carry out periodic bench cycle test（e.g.WHSC/WHTC）,the NO_x emission deterioration equation used to calculate the NO_x emission deterioration factor was fitted according to the test results.The NO_x emission in-use compliance detection of the heavy-duty vehicles should be conducted by PEMS test in China Ⅵ standard.In order to expand the NO_x emission deterioration and in-use compliance research method,meanwhile,considering the China Ⅵ（B）standard requirements for data uploading in the whole life cycle of heavy-duty diesel vehicles,this paper intended to conduct the deterioration analysis of engine out NO_x emission and SCR catalyst,and NO_x emission in-use compliance analysis of a N3 catagory vehicle by using on-board NO_x sensor data,engine operation data,ambient data,etc.collecting from 250000 km operation（lasted for 13 months,the ambient temperature ranged from-4 to 42℃,the relative humidity ranged from 3.2 to 99.8%）and in-use compliance test and uploading to the cloud platform.In this paper,the deterioration analysis and its prediction methods of engine out NO_x emission and the deterioration analysis method of SCR catalyst using on-board NO_x sensor data was proposed,the method of using on-board NO_x sensor data to check the NO_x emission in-use compliance was established,the classification method of the valid moving average windows and the NO_x emission in-use compliance evaluation rule was analyzed,the suggestion of optimizing the evaluation rule was put forward.The main research contents and conclusions of this paper are as follows:（1）First of all,the clean dataset was obtained by a series of pre-processing including timing check,outlier processing and derivative parameter calculation,etc.Then the engine operation parameters,intake parameters,fuel parameters,ambient parameters were taken as the constraint conditions successively to get a same steady dataset for engine out NO_x deterioration analysis,the formula for on-board NO_x correction for ambient temperature and humidity was fitted on the steady dataset with maximum weight.Finally,steady weighted conditions during the operation of the tested vehicle were extracted,after NO_x correction for ambient temperature and humidity,the engine NO_x emission deterioration equation was fitted by the weighted brake specific engine out NO_x emission of steady weighted conditions in different mileage ranges with 10000 km as the step.（2）Different machine learning models were trained on the dataset composed by0～10000 km（the shortest mileage for in-use compliance detection）on-road operation data of the tested vehicle.It was found and verified that the random forest model had good generalization ability for engine out NO_x emission prediction.In different mileage ranges within 250000 km,engine out NO_x emission deterioration factors predicted by random forest model were not more than 1.091,the error was within 9%compared to the corresponding result of steady-state.（3）The percentage change of engine torque,the change rate of engine out and tailpipe NO_x concentration,the average temperature of SCR catalyst,exhaust mass flow and ambient parameters as constraint conditions to explore the method of extracting the steady weight dataset.Then the abnormal points of NO_x conversion efficiency caused by ammonia nitrogen ratio（SCR control strategy）or NH₃ slip should be eliminated as far as possible so as to obtain the dataset that could be used for SCR catalyst deterioration analysis on the steady weight dataset.The deterioration state of SCR catalyst was determined by calculating the average NO_x conversion efficiency and average ammonia nitrogen ratio in different mileage ranges with 10000 km as the step.（4）The NO_x emission in-use compliance detection method based on on-board NO_x sensor data was established.According to this method,12 different load tests were carried out during the operation of the tested vehicle.The results showed that the NO_x emissions obtained based on on-board NO_x sensor data were all higher than the corresponding results obtained by PEMS.The deviations between the two mainly ranged from 1.14%to 11.35%.Therefore,if the NO_x emission based on the NO_x sensor met the in-use compliance requirement,the corresponding results measured by PEMS could also met the requirement.（5）The analysis results of 12 different payloads in-use compliance tests showed that the valid moving average windows obtained based on China Ⅵ standard were all composed by the data generated during the rural and motorway section,that means the data generated during the urban section was not included in the dataset for the NO_x emission evaluation.The NO_x mass emission from urban section accounted for a high proportion in the whole test,leading to NO_x emissions in-use compliance evaluation results were significantly lower than the real NO_x emission.In view of this,the power threshold,definition rule and weight setting of the valid moving average windows were repeatedly adjusted and checked,and finally the evaluation rule which could more truly characterize the on-road real NO_x emission level of the tested vehicle was proposed.