A novel optimization-based method to develop representative driving cycle in various driving conditions

The lack representativeness of in-used driving cycles has raised substantial concerns regarding the enlarging gap between real-world fuel consumption and type-approval. Considering the high randomness of existing driving cycle development methods, the developed cycle still has low representativeness in capturing the patterns in the real-world. In this study, a novel data-driven driving cycle development method MMACO-MC based on Min-Max Ant Colony Optimization (MMACO) and Markov Chain is proposed to improve the representativeness of driving cycles. The proposed MMACO-MC is then applied to develop driving cycles in Fuzhou city under various driving conditions. Significant differences in cycle parameters have been observed in different driving conditions, which further lead to a 15% deviation on the FCR estimation (Fuel Consumption Rate). Meanwhile, the FCR estimation in the whole region of Fuzhou also deviates from the standard cycles from 22.8% to 29.4%. Lastly, the optimal cycle length is explored to ensure the stability of FCR estimation under various traffic scenarios. This study highlighted the necessity of optimization-based driving cycle development in the accuracy of fuel consumption estimation. The proposed method and the conclusions could be applied as a reference by the authorities to establish fuel consumption standards in the future.(c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

The lack of representativeness in existing driving cycles has led to concerns about the increasing gap between real-world fuel consumption and type-approval. This study proposes a novel data-driven driving cycle development method to improve the representativeness of driving cycles. The research finds significant differences in cycle parameters under different driving conditions, resulting in a large deviation in fuel consumption rate estimation.