论文部分内容阅读
The vehicle specific power(VSP)distribution is an important parameter in emission modelling,like the Motor Vehicle Emission Simulator(MOVES).Using the default VSP distributions in MOVES to calculate local emissions may cause error.It is recommended to collect local vehicle activity data to develop specific VSP distributions.It raises the problem that how much data are needed.This study investigates the sample size for developing robust VSP distribution on expressway for emissions estimation.The Root Mean Square Error(RMSE)is employed to quantify the uncertainty of developing VSP distribution.The case study indicates that the error of VSP distribution drops significantly when increasing data sample size.After the sample reaches 420 seconds,the error maintains at a low level.The emissions estimation error drops from 23%to 4.8%when increasing sample size from 60 seconds to 420 seconds.