关键词: 履带式车辆, 循环工况, 马尔科夫链, 特征提取, 车辆动力学

Abstract: To evaluate the performance of tracked vehicles, a multi-road type, three-dimensional driving cycle construction method based on micro-motion segments is proposed. This method aims to address the issues of variety of road types, lengthy short-trip segments, and the numerous dimensions of influencing factors in the construction of driving cycles for tracked vehicles. The collected driving data is cleaned, and three-dimensional data such as speed, angular velocity, and ground resistance coefficient are extracted. Using the k-means clustering method, the driving segments are categorized into three typical road surfaces: paved road, gravel road, and undulating dirt road. Based on the minimum weighted three-dimensional variation rate of the short-trip segments, they are divided into micro-motion segments, followed by feature extraction and clustering analysis. A three-dimensional driving cycle is constructed using the Markov transition probability of the micro-motion segments, and a corresponding comprehensive evaluation system is proposed. The total duration of the constructed driving cycle is approximately 2000 seconds, with an average feature coverage rate of 94.63% across the three road types. This driving cycle accurately reflects the driving characteristics of tracked vehicles and serves as an effective tool for simulation and bench testing of tracked vehicles.

Key words: tracked vehicles, driving cycle, Markov chain, feature extraction, vehicle dynamics