基于模糊推理的驾驶员换道模型

Lane-changing Models Research Based on Fuzzy Inference

以换道过程中目标车道跟随车为研究对象,对跟随车与换道车之间的交互行为进行分析,采用模糊推理技术进行建模;选取相对间距、相对速度、最迟换道距离、驾驶员性格等作为模糊推理系统输入变量,换道支持度为输出,构造3类不同换道方式的模糊规则,建立基于模糊推理的车辆换道模型。结果表明:根据模糊推理的特点设计的换道模型能够反映车辆驾驶行为的自主特性;通过改变交通流密度进行数值模拟分析发现,相比于强制换道和对称双车道元胞自动机模型,协作换道模糊推理模型提高了整个路段的交通流平均速度,减少了路段交通拥挤。

The interaction between the following and the lane-changing vehicle was analyzed regarding the following vehi- cle of the target lane as the research object during lane-changing process. Fuzzy inference technique was used to proceed with modeling studies. Relative velocity, expected relative distance, latest lane change distance, and driver characteris- tics were selected as the fuzzy inference system input variables, and lane change support as the output. The fuzzy rules of three lane-changing ways were extracted and the lane-changing models were built based on fuzzy reasoning. According to these features of fuzzy control, this method reflects the autonomous characters of vehicle driving behaviors. By changing traffic flow density, the results show that the cooperate lane-changing improves the average speed of whole road and re- duces the road traffic congestion, comparing with the forced and symmetric two-lane cellular automata.