Prediction Model Of Vehicle Exhaust Emissions At Expressway Toll Stations

This study evaluates the vehicles' emission characteristics and traffic flow characteristics for two kinds of highway toll collection lanes(electronic toll collection,ETC and manual toll collection,MTC)in China.Statistics analysis methods,queuing theory,and mathematical modeling algorithm are adopted to measure and estimate vehicles' exhausts,including CO,HC,NOx,and CO2.The emission estimation models of CO,HC,NOx,and CO2 are constructed by two different methods,and model comparison and evaluation results are obtained.The traffic flow characteristics and the real-time emission data of CO,HC,NOx,CO2 were collected with unmanned aerial vehicle(UAV)and AUTOplus,respectively.The main contents of this study are listed as follows.(1)Four vehicle types are classified,i.e.passenger cars,buses,vans,and heavy trucks.Then the vehicular traffic flow characteristics(including speed,acceleration,volume,delay,and queue length)during peak and off-peak periods were calculated.Following that,comparason analysis and significance tests were performed and the results indicate that with the help of ETC system,traffic efficiency and the LOS of highway toll stations could be improved significantly due to smaller influence area,less delays and shorter queues of the toll lanes equipped with ETC system.(2)To measure the emission characteristics of different vehicle types in ETC and MTC lanes,real-time mass emission rates collected by AUTOplus and the real-time speed as well as acceleration data collected via portable GPS are fused and matched.Then the emission rules were figured out from three research levels,which including emission rules of vehicles in different working conditions,spatial and temporal variety regulation of emissions and the influence of traffic parameters(speed and acceleration)on emissions.Compared with MTC lanes,ETC lanes are characterized by short variable speed and distance range(be shortened by 40-60m),low total emissions(be reduced by at least half for different gases),and uniform spatial distributions of emissions.Besides,the exhaust emission level in the high-speed intervals is higher than that in the low speed range basically,while the emission level in the high acceleration intervals is gennerally higher than that in the low acceleration intervals.What's more,with the increase of acceleration,the emission factors increased with fluctions in ETC lanes,and there were no significant emissions at the zero acceleration point.Finally,there were obvious emission concentrations duiring the idling mode of vehicles in MTC lanes.(3)The vehicle specific power(VSP)index was introduced to establish real-time emission forecasting models and emission quantity forecasting models in highway toll stations.Compared with traditional VT-Micro emission models building the relationship between emissions and speed as well as acceleration,the regression models by quadratic polynomial fitting between average mass emission rates and VSP are more accurate to measure the emission levels of CO,HC,NOx and CO2.Moreover,with the increase of VSP,emissions increase with fluctuations in both ETC and MTC lanes.This indicates that vehicles would produce more exhausts during idling and acceleration modes.