The Research Of Emission Estimation Of Light-duty Gasoline Vehicles Based On Vehicular Networking

As one of the major contributors to air pollution which becomes increasingly serious,motor vehicles are forced to comply with the more and more stringent emission regulations announced by the government.However the regulatory cycle can not accurately reflect the level of emissions of the vehicles in the actual road.The current regulations require that the On-Board Diagnostic system be able to accurately alert the exhaust system’s problem,but it can not force the driver to deal with it,resulting in a large number of vehicles with emission problems traveling on the road.Therefore,it is necessary to develop a remote monitoring system.In order to achieve effective monitoring,it should not only be able to obtain vehicle status by remote diagnosis,but also can estimate real-time exhaust based on the monitoring data.First of all,the methodology of emissions estimation is studied.One is to determine the key factors affecting the level of the vehicles’ emissions in the actual road driving process.By means of simulation experiment,the driving conditions,speed,acceleration,load and slope are respectively set,and the qualitative analysis of the result is maked.And summed up the three key factors of the speed,acceleration and power.The second is to determine an effective method to construct the vehicle emission estimation model.Based on the study of neural network,the model is established by radial basis function neural network method,and the lack of radial basis function neural network is improved.Improvement includes determining the number of hidden nodes by the Resource Allocation Network(RAN)algorithm;simplifying the network by the pruning strategy;global optimal searching of the network parameters by the improved Particle Swarm Optimization(MPSO)algorithm.Taking the simulation experiment data as the sample,the constructed model works well.Secondly,a data acquisition system based on vehicular networking is designed and developed.It mainly includs the vehicle terminal and the back-end server-side.The vehicle terminal contains CAN interface,2G module,GPS module,power module and start the detection module,and STM32 as the main chip,achieving data acquisition,data upload and power management functions.Set up a data management center,including servers,servers,and clients,to meet the requirements for receiving uploaded data,storing data,and querying data.Finally,the actual road emission test experiment of motor vehicle is used to verify the emission estimation method and the data acquisition system.The exhaust emission data and status data of the vehicle running on a reasonably pre-selected route are collected simultaneously by a Portable Emission Measurement System(PEMS)and the data acquisition system.In this paper,estimation of vehicle emissions on the actual road were studied in depth,including simulation analysis,model establishment,system development and experimental verification.And finally real-time monitoring of vehicle emissions is realized.The achievement not only can strengthen the supervision of vehicle emissions,but also can be used as a basis for traffic management to control the total regional emissions.