面向商用车行驶工况优化设计的高速公路工况识别

为解决商用车行驶工况优化设计中确定工况类型的问题,研究商用车行驶工况特性,提出一种基于朴素贝叶斯方法的高速公路工况识别方法。利用21辆长途运营商用车采集的106 200 km行驶工况数据,以3 km为单位进行分割,共获得35 230段有效试验路段数据(其中:高速公路27 986段;一般公路6 124段;城市公路1 120段)。以该数据为基础,根据朴素贝叶斯方法分析汽车运行过程中的平均速度和挡位统计信息,确定面向商用车行驶工况优化设计的阈值划分区间,获得相关的先验概率和条件概率,利用MATLAB软件进行编程计算,对高速公路工况进行了识别分析。研究结果表明:高速公路工况识别的正确率到达88.26%,高速公路工况被误判为一般公路工况的误判率为9.54%,高速公路工况被误判为城市公路工况的误判率为2.20%;基于朴素贝叶斯方法的高速公路工况识别能够为商用车行驶工况优化设计提供一种有效的高速公路工况识别方法。

Identification of Expressway Driving Cycles for Optimization of Commercial Vehicle Driving Cycles

To solve the identification problem of the driving cycles type for optimization of commercial vehicle driving cycles, the characteristics of commercial vehicle driving cycles were analyzed. An expressway driving cycles identification method based on the Naive Bayes method is proposed. The 106 200 km data collected by 21 long-distance commercial vehicles were divided into individual 3 km segments, and 35 230 effective experimental road segments (including 27 986, 6 124, and 1 120 segments of expressways, general highways, and urban roads, respectively) were obtained. Based on the above data, according to the Naive Bayes method, the average speed and gear statistical information of the commercial vehicle driving cycles were analyzed, the threshold division intervals for optimization of commercial vehicle driving cycles were determined, and the relevant prior probability and conditional probability were obtained. The types of driving cycles were discriminated and analyzed by programming calculations using MATLAB software. The results indicate the following:the identification accuracy of expressway driving cycles reaches 88.26%, the misjudgment of expressway driving cycles as general highway driving cycles is 9.54%, and the misjudgment of expressway driving cycles as urban highway driving cycles is 2.20%. The identification method of expressway driving cycles based on the Naive Bayes method is effective for identifying expressway driving cycles for optimization of commercial vehicle driving cycles.