基于贝叶斯网络的高速公路出口指路标志满意度评价

Satisfaction Evaluation of the Setting of Guide Signs at Expressway Exits Based on Bayesian Networks

为建立科学有效的高速公路出口指路标志设置满意度评价体系,分析多维度因子对评价结果的直接和潜在影响。采用问卷调查收集驾驶人的基本信息和对指路标志的主观评价,通过贝叶斯网络构建高速公路出口指路标志设置满意度评价模型。结果表明,文中建立的贝叶斯网络模型可以准确识别驾驶人对高速公路出口指路标志的满意度,总体准确率达到72.9%;通过贝叶斯网络模型识别了驾驶人个体特征、出行选择特征、出口处的行为特征等因素之间的因果关联及其组合对出口指路标志满意度的影响。

In order to establish a scientific and practical satisfaction evaluation system for the setting of guide signs at expressway exits,the direct and potential impacts of multi-dimensional factors on the evaluation results are analyzed.The questionnaire survey is used to collect basic information about drivers and the subjective evaluation of guide signs.The satisfaction evaluation model of the setting of guide signs at expressway exits is constructed through a Bayesian network.The results show that the proposed Bayesian Networks model can accurately identify drivers’satisfaction of the guide signs at the expressway exits,with an overall accuracy rate of 72.9%.In addition,the Bayesian Networks model is used to identify the causal relationships between drivers’individual features,travel choice characteristics,and behavior characteristics at exits,as well as the impact of their combinations on the satisfaction of exit guide signs.