Trajectory planning for autonomous modular vehicle docking and autonomous vehicle platooning operations
Trajectory planning for autonomous modular vehicle docking and autonomous vehicle platooning operations
复制标题
DOI：
10.1016/j.tre.2022.102886
复制
发表时间：
2022-10
期刊：
影响因子：
--
通讯作者：
Qianwen Li;Xiaopeng Li
中科院分区：
文献类型：
--
作者：
Qianwen Li;Xiaopeng Li
新兴的自主模块化车辆（AMV）技术允许车辆单元在途中物理对接或彼此分离，以形成不同长度的车辆。该技术在道路物流中具有巨大的潜力，其中形成排/长列车来运输货物和乘客，即，货运和运输系统。AMV对接是自动车辆（AV）排队的极端情况，因为AMV以零间隙物理连接。本文将AMV对接和AV编队轨迹规划问题转化为两阶段优化问题。提出了可行锥方法，揭示了解的可行性的理论性质，并解析地求解了第一阶段问题。这种方法提供了一个简约的启发式方法来设计指定为几个二次段的轨迹的基础。基于庞特里亚金最大值原理，提出了一种启发式的替代解决方案，以解决原问题的一个特殊情况下的精确最优。然后提出了一种基于二次规划的精确求解方法来优化轨迹。采用可行锥方法构造有效割集，提高了精确求解的效率。数值实验表明，与精确解方法相比，简约启发式方法能获得近似最优解，大大缩短求解时间，适合于实时工程应用。结果还表明，在优化AMV对接和AV编队轨迹与传统的编队方法相比，简约启发式方法的优越性。敏感性分析结果揭示了建议参数选择的排相关的物流，以平衡之间的权衡运作效率和成本。
Emerging autonomous modular vehicle (AMV) technology allows vehicle units to physically dock on or split from each other en route to form vehicles of different lengths. This technology has great potential in roadway logistics where platoons/long trains are formed to transport goods and passengers, i.e., freight and transit systems. AMV docking is an extreme case of autonomous vehicle (AV) platooning in that AMVs are physically connected with zero gaps. This paper formulates the AMV docking and AV platooning trajectory planning problem into a two-stage optimization problem. A feasible cone method is proposed to reveal the theoretical properties of solution feasibility and solve the first-stage problem analytically. This method provides the basics for a parsimonious heuristic approach to design trajectories specified as several quadratic segments. A heuristic alternative solution based on Pontryagin's maximum principle is proposed to solve a special case of the original problem to the exact optimum. Then an exact solution approach based on quadratic programming is proposed to optimize the trajectories. The feasible cone method is used to construct valid cuts to expedite the exact solution efficiency. Numerical experiments show that the parsimonious heuristic approach can achieve near-optimal solutions and greatly reduce the solution time compared with the exact solution approach, appealing to real-time engineering applications. The results also demonstrate the superiority of the parsimonious heuristic approach in optimizing AMV docking and AV platooning trajectories compared with traditional platooning methods. Sensitivity analysis results shed insights into advising parameter selections of platoon-related logistics to balance the tradeoff between operational efficiency and cost.