摘要： 采集黄灯期间驾驶员行为的相关数据，考虑车辆位置、车速、倒计时表3个影响因素，分别设定其隶属度函数，应用模糊决策树中的FID3算法，以模糊信息熵为启发信息，构建驾驶员选择的模糊决策树模型，生成决策规则。利用测试样本对模型进行检验，结果表明，基于模糊决策树的预测结果准确率总体达到84.8%。

关键词: 数据挖掘, 驾驶员行为, 模糊决策树, 不确定理论, 模糊信息熵

Abstract: Drivers decision to go or stop during the yellow interval belongs to uncertain decision making. This paper collects drivers behavior data at four similar intersections. Fuzzy Decision Tree(FDT) is applied to model driver behavior at signalized intersection. Considering vehicle location, velocity and countdown timer as the influencing factors, the FDT model is constructed using FID3 algorithm, and decision rules are generated as well. Test sample is applied to test FDT model, and results indicate that FDT model can predict drivers’ decision with overall accuracy of 84.8%.

Key words: data mining, driver behavior, Fuzzy Decision Tree(FDT), uncertain theory, fuzzy information entropy

中图分类号: 

• TP273+.4