Development and performance of a new vehicle emissions and fuel consumption software (PΔP) with a high resolution in time and space

Developmental Biology

Robin Smit¹ 

[1]Transport Emission Research, 16 Mayleen Street, Clontarf, QLD 4019, Australia$$

关键词: Road transport;    emission;    fuel consumption;    high resolution;    road traffic;   

DOI  :  10.5094/APR.2013.038

学科分类：农业科学（综合）

来源: Dokuz Eylul Universitesi * Department of Environmental Engineering

【 摘 要 】

This paper reports on the development and performance of a new simulation tool for road vehicles. The PΔP model predicts second-by-second fuel consumption, air pollution (NO_X) and greenhouse gas emissions (CO₂) with a high resolution in time and space. It uses engine power and the change in engine power as the main model variables to simulate vehicle fuel consumption and emissions for all relevant vehicle classes including cars, SUVs, light–commercial vehicles, rigid trucks and articulated trucks. A total of 73 vehicle classes are modeled accounting for main vehicle type, fuel type and technology level. The model uses data from a large verified Australian emissions database containing around 2 500 modal emission tests (1 Hz) and about 12 500 individual bag measurements. The minimum input requirements for the model are speed-time data (1 Hz) and vehicle types. This kind of information is typically available from microscopic traffic simulation models, on–road measurements or expert judgment. The user of the model can also specify the road gradient, the vehicle loading and the use of air conditioning. Default values are provided for each of these where location-specific data are unavailable. The PΔP model aims for an optimum balance between model complexity and prediction accuracy. The performance results for the PΔP model results are good with, for instance, average R² values of 0.65 and 0.93 for NO_X and CO₂/fuel consumption, respectively. This performance compares well with that reported for other models with different complexity. The emission algorithms are shown to be robust with respect to prediction errors. Aggregation of the 1 Hz prediction results in time/space (e.g. 100 m road segments) and across vehicle classes (e.g. passenger car, SUV, articulated truck, etc.) further improves prediction performance.

【 授权许可】

Unknown   

【 预 览 】

附件列表

Files	Size	Format	View
RO201912040527736ZK.pdf	679KB	PDF