Construction of Battery Electric Vehicle Driving Cycles Based on Improved Grey Wolf Optimized K-Means Clustering: A Case Study for Qingdao, China

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Standardized driving cycles often inadequately represent the driving patterns specific to a particular city, and variations in vehicle types within the same city further contribute to discrepancies in driving cycles. This study seeks to characterize the driving patterns of a specific vehicle model within a designated city and to provide robust data support for precise predictions of energy consumption and driving range. To achieve this objective, a driving cycle was developed and analyzed using real-world operational data collected from a battery electric vehicle (BEV) in Qingdao, China. The driving cycle was constructed through a process involving data preprocessing, dimensionality reduction via principal component analysis (PCA), and Improved Grey Wolf Optimizer K-means (IGWO-K-means). The analysis of energy consumption per 100 km is concluded by the study. Validation of the constructed driving cycle against the preprocessed data yielded an average relative error of 2.31%, providing a reference for the real-world driving cycle of BEVs in Qingdao, China. Furthermore, a comparative analysis of the driving cycles for BEVs in Qingdao, China, and internal combustion engine vehicles (ICEVs) in Fuzhou and Nanjing, China, revealed notable differences. This underscores the critical need for developing driving cycles that are specifically tailored to distinct cities and vehicle models. The examination of energy consumption per 100 km further corroborated the representativeness of the constructed driving cycle. Furthermore, a comparative assessment of energy consumption across varying ambient temperature ranges demonstrated that it increases as temperatures decrease.

This publication has 22 references indexed in Scilit: