关键词: 行驶工况；主成分分析；聚类分析；公交线路；本地化建模

Abstract: To address the issue of insufficient matching between current vehicle drive cycle standards and the actual traffic operation characteristics of local areas, this study focuses on a typical bus route in Taiyuan (route 849) and proposes a method for constructing localized drive cycle by considering road segment complexity (congested sections, slopes, etc.). Actual operational data are collected using GPSpro equipment, and based on the short-trip segmentation theory, 529 valid kinematic segments are extracted. Fifteen characteristic parameters, including average speed and acceleration standard deviation, are used to construct a feature matrix. Principal component analysis is applied for dimensionality reduction of the features, and the K-means clustering algorithm is used to classify the segments into three typical driving patterns: congested, normal, and smooth. Finally, by combining 12 representative short trips, a localized drive cycle curve conforming to the operational characteristics of urban buses in Taiyuan is constructed. Verification shows that the relative errors of various characteristic parameters between the constructed drive cycle and the original test data are all below 8%, indicating that this method can effectively reflect the real operational status of regional buses and provide data support for vehicle selection, energy consumption assessment, and route optimization.

Key words: drive cycles; principal component analysis; cluster analysis; bus routes; localization modeling