结合自然语言处理与改进层次分析法的乘用车驾驶舒适性评价

Driving comfort evaluation of passenger vehicles with natural language processing and improved AHP

为了有效和准确地评价乘用车驾驶舒适性,构建了"人-机-环境"系统下的汽车驾驶舒适性本体模型,并对驾驶舒适性的形成机制进行了阐述。通过用户访谈实验并基于改进的"词频-逆向文件频率"(TF-IDF)方法对主题词进行提取,获得影响驾驶舒适性的主要因素并进行分类,在此基础上建立递阶层次结构;针对传统层次分析法(AHP)的不足,以三标度取代九标度以保证精确性,通过Delphi实验,构造两两比较判别矩阵后再进行处理,以计算各评价指标的权重,并对该指标体系进行一致性检验。以C级乘用车为例,验证了该评价方法的有效性。该方法为乘用车的驾驶舒适性评价提供了一种可行的技术支持。

The driving comfort of drivers in passenger vehicles was evaluated using an ontology model for driving comfort in a human-machine-environment system.The model was then used to evaluate driving comfort with the main driving comfort description extracted and classified via user interviews based on the improved term frequency-inverse document frequency（TF-IDF）method.The words were then ranked in a hierarchical structure.The scale of nine in the traditional analytical hierarchy process（AHP）was replaced by a scale of three to improve the accuracy.The comparative judgment matrix was defined with the weight of each index calculated through a Delphi survey.The index consistency system was also tested.The reliability of this method was validated using C-class passenger vehicles. Thus,this gives an effective approach for evaluating vehicle driving comfort.