基于决策树模型的驾驶员期望车速  

Driver's desired speed based on decision tree model

作　　者：王晓原[1] 邢丽[1] 吴芳[1] 

机构地区：[1]山东理工大学交通与车辆工程学院,山东淄博255049

出　　处：《计算机应用》2009年第B12期318-321,共4页journal of Computer Applications

基　　金：山东省自然科学基金资助项目(Y2006G32);山东理工大学科研基金重点资助项目(2004KJZ02)

摘　　要：微观交通流仿真是智能运输系统(ITS)研究与开发的重要手段。期望车速是微观交通流仿真研究的一个重要参数,受到驾驶员特性、车辆特性、道路条件、交通干扰、天气和承运任务急缓等多种因素的影响,准确地确定期望车速是驾驶员行为研究的难点。从研究驾驶员心理—物理特性的角度出发,利用决策树能融知识表示与获取于一身的优点,将决策树用于驾驶员期望车速的研究,以实现对驾驶员行为的模拟再现。仿真结果表明,该方法用于驾驶员期望车速的研究是可行的。Microcosmic traffic simulation is an important means for studying theIntelligent Transportation System （ITS）. The desired speed is an important parameter in microcosmic traffic simulation. The desired speed is influenced by many different factors, such as the driver characteristics, the vehicle characteristics, the road condition, the traffic interference, weather and the emergency of assignment. It is difficult to determine the desired speed exactly for studying driving behavior. Based on the driver＇s psycho-physic characteristic and the fact that the knowledge representation and acquisition could be fused, decision tree was used to study the driver＇s desired speed in order to simulate driving behavior. Simulation results show that the algorithm is feasible.

关 键 词：交通工程 驾驶员行为 决策树 期望车速 

分 类 号：U491[交通运输工程—交通运输规划与管理] TP18[交通运输工程—道路与铁道工程]