Driving cycles are widely used as standard measurement procedures for vehicle emissions and energy consumption assessment, especially for vehicle design optimization and performance evaluation of emerging vehicular technologies (HEV/EV), including standard driving cycles and non-standard driving cycles. Local driving cycles, as non-standard driving cycles, are mainly used in certain areas for in-use vehicle emission and fuel consumption assessment. However, whether these local driving cycles could represent the driving conditions in certain areas for a long time, and whether it is necessary to consider the year of development when comparing local driving cycles in different cities, these issues require more attention from researchers. In this paper, a practical local driving cycle construction methodology combing principal components analysis (PCA) and K-means clustering is proposed, and two local driving cycles for passenger cars in Nanjing are developed in 2009 and 2017, respectively. The same data acquisition and processing methods make the comparison between the two generated cycles more fair. Significant differences in driving parameters between Nanjing local driving cycles in 2009 and 2017 show the necessity of studying local driving cycles from time dimension and provide traffic management authority with some valuable advice on estimating road-based vehicle emission.
Read full abstract