Representative driving cycle development and fuel economy evaluation: Case studies in China

Abstract

With the increasing adoption of hybrid electric vehicles (HEVs), driving cycles have become essential tools for evaluating their fuel economy. However, existing standard driving cycles often fail to account for significant variations in topography and traffic patterns among different cities in China. To address this gap, this study collected over 200,000 km of real-world driving data from nine major Chinese cities, including Shanghai, Shenzhen, and Chongqing. A representative driving cycle generation algorithm based on the chi-square test of characteristic distribution was developed, producing driving cycles tailored to each city. These cycles were used to evaluate the fuel economy of a specific HEV on a Hardware-in-the-Loop (HIL) platform. The results show that the generated driving cycles maintain durations of 30–40 min, with p-values from the chi-square goodness-of-fit test exceeding 0.05, indicating no significant differences between the observed and expected distributions. Fuel economy evaluations revealed a negative correlation (−0.805) between average speed and fuel consumption, and a positive correlation (0.827) between average gradient and fuel consumption. These findings demonstrate that the developed driving cycles are both representative and effective for assessing HEV performance under diverse urban conditions.