高速公路浓雾环境下换道决策规则提取及决策算法

依据换道决策规则进行换道是当前无人驾驶车辆常用的决策方法之一。针对浓雾环境下换道决策规则提取困难和研究较少的问题,研究了高速公路浓雾环境下的换道决策行为。首先,招募24名职业司机,利用Auto Sim驾驶模拟舱搭建虚拟高速公路浓雾环境进行驾驶实验;其次,提出了基于CART决策树的换道决策规则提取方法,提取出15条换道决策规则;最后,对换道决策规则进行了验证。结果表明,用CART决策树算法提取高速公路浓雾环境下换道决策规则是可行的,提取的规则能准确反应驾驶员换道行为的决策过程,可为高速公路浓雾环境下无人驾驶车辆的换道决策提供一定的理论支撑。

Decision-making Rule Extraction and Decision-making Algorithm for Lane Change in Dense Fog Environment

Changing lanes according to the lane change decision rules is one of the common decision-making methods for unmanned vehicles. In view of the difficulty in extracting decision rules and the less research in the dense fog environment, this paper studies the decision-making behavior of lane change in dense fog environment of expressway. First, 24 professional drivers were recruited for the experiment, and the AutoSim driving simulator was used to build a virtual highway dense fog environment for driving experiments; Second, a method for extracting lane change decision rules based on CART decision tree was proposed, and 15 lane change decision rules were extracted from it; Final, the lane change decision rules were verified. The results show that it is feasible to use the CART decision tree algorithm to extract the lane change decision rules in the dense fog environment of the expressway; The extracted rules can accurately reflect the decision-making process of the driver''s lane change behavior, which can provide a certain theoretical support for the lane-changing decision-making of unmanned vehicles in dense fog environment of expressway.