Lane change decision planning for autonomous vehicles
Lane change decision planning for autonomous vehicles
复制标题
自动驾驶车辆变道决策规划
DOI：
--
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
Yunhai Zhu
中科院分区：
文献类型：
--
作者：
Kangqiang Ouyang;Yong Wang;Yanqiang Li;Yunhai Zhu
本文构建了一个简化的换道场景模型。在实际驾驶场景中，频繁换道会降低驾驶的安全性和舒适性。本文利用增益函数分析换道意图的产生，并依靠最小安全距离公式来判断换道风险。当满足相关条件时，车辆执行换道动作，否则继续保持跟车状态。本文基于五次多项式的一般模型生成换道轨迹，并通过计算损失函数的值来选择相对最优的换道轨迹。最后，通过MATLAB仿真环境进行验证和分析。结果表明，本文提出的方法在换道动作的安全性和稳定性方面有了显著提高。
This article constructs a simplified lane-changing scene model. Frequent lane-changing in actual driving scenarios will reduce driving safety and comfort. This article uses a gain function to analyze the generation of lane-changing intentions and rely on the minimum safety distance formula to judge The risk of changing lanes. When the relevant conditions are met, the vehicle executes the lane-changing action, otherwise it continues to maintain the car following state. In this paper, the lane-changing trajectory is generated based on the general model of the fifth degree polynomial, and the relatively optimal lane-changing trajectory is selected by calculating the value of the loss function. Finally, it is verified and analyzed through the MATLAB simulation environment. The results show that the method proposed in this paper has achieved a significant improvement in the safety and stability of the lane changing action.