A Monte Carlo simulation methodology is suggested in order to assess the impact of ambient wind on a vehicle’s performance and emissions. A large number of random wind profiles is generated by implementing the Weibull and uniform statistical distributions for wind speed and direction, respectively. Wind speed data are drawn from eight cities across Europe. The vehicle considered is a diesel-powered, turbocharged, light-commercial vehicle and the baseline trip is the worldwide harmonized light-duty vehicles WLTC cycle. A detailed engine-mapping approach is used as the basis for the results, complemented with experimentally derived correction coefficients to account for engine transients. The properties of interest are (engine-out) NO and soot emissions, as well as fuel and energy consumption and CO2 emissions. Results from this study show that there is an aggregate increase in all properties, vis-à-vis the reference case (i.e., zero wind), if ambient wind is to be accounted for in road load calculation. Mean wind speeds for the different sites examined range from 14.6 km/h to 24.2 km/h. The average increase in the properties studied, across all sites, ranges from 0.22% up to 2.52% depending on the trip and the property (CO2, soot, NO, energy consumption) examined. Based on individual trip assessment, it was found that especially at high vehicle speeds where wind drag becomes the major road load force, CO2 emissions may increase by 28%, NO emissions by 22%, and soot emissions by 13% in the presence of strong headwinds. Moreover, it is demonstrated that the adverse effect of headwinds far exceeds the positive effect of tailwinds, thus explaining the overall increase in fuel/energy consumption as well as emissions, while also highlighting the shortcomings of the current certification procedure, which neglects ambient wind effects.
Recurrent Neural Network Based Driving Cycle Development for Light Duty Vehicles in Beijing
