Abstract:Heavy-duty commercial vehicles are usually customized to meet the specific demands of different market segments. This paper proposes an innovative research approach to construct driving cycles for different market segments of heavy-duty commercial vehicles. To verify the rationality of this research approach, the study takes the daily industrial product market segments as a research case. Relying on the onboard Big-data system to collect users’ driving data from 3,000 National VI series semi-trailer trucks in this market. Through a series of steps, including kinematic segmentation, data dimensionality reduction, and segment chaining, three representative driving cycles were constructed. Based on this, a simulation model was built in AVL Cruise to predict the fuel consumption of the target market users and compare it with the actual fuel consumption and the predicted results based on legislative standard driving cycles (C-WTVC and CHTC-TT). The results show that compared with C-WTVC and CHTC-TT, the constructed driving cycles for the industrial market segment are closer to the real-world driving characteristics of this market, with an average relative error reduction of 18.67% and 32.97% for the characteristic parameters, respectively. It can also more accurately predict users’ fuel consumption, with an improvement in prediction accuracy of 7% and 4%, respectively. Therefore, constructing driving cycles for the submarkets of heavy-duty commercial vehicles can more accurately characterize the vehicle driving patterns of target users, thus improving the prediction accuracy of fuel consumption. |