基于轨迹数据的货车自发编队节油潜力估计

为探索最大编队长度和路网中行驶货车数目对自发编队的影响以及评估真实世界自发编队油耗节省潜力，提出了既可保证一定程度的搜索又能防止为过分追求更大编队长度而花费较长时间等待的动态时空搜索范围策略，利用辽宁省货车轨迹数据挖掘了多辆货车的最长公共子序列，并构建整数规划模型求解，获取最大节省油耗下的编队方案。结果显示：编队油耗节省会随着路网中行驶货车数目和最大编队长度的增加而增加，编队中车辆平均行驶距离和平均节省油耗最终分别收敛至一范围；编队行驶距离并不一定会随着允许最大编队长度的增加而增加，即自发编队最大油耗节省策略并非编队最大行驶距离策略；在不考虑时间搜索范围时，通过继续增大空间搜索范围可实现更多的燃油节省，但不考虑空间搜索范围时持续增加时间搜索范围却不能获得更多燃油节省；4000辆车的数据集进行自发编队时每天最多可节省燃油2026.21 L，最大允许编队长度为5辆比最大允许编队长度为2辆的车辆平均节省油耗最多可提升13.92%。将本文整套处理流程应用于更大规模的货车数据集，可组成自发编队的目标车辆将会显著增加，相应地也可减少更多的能源消耗和污染物的排放，具有很大的应用潜力。

Estimating Truck Spontaneous Platoon Fuel-saving Potential Based on Trajectory Data

This study explores the influence of the maximum truck platoon size and the number of trucks on a spontaneous platoon and evaluates the fuel-saving potential in the real world. A dynamic spatiotemporal search scope was proposed to ensure a certain degree of exploration and prevent vehicles from spending much waiting time caused by the excessive pursuit of a larger-sized platoon. Moreover, we used the truck trajectory data of Liaoning Province to mine the Longest Common Subsequence (LCSS) of trucks, and an integer programming model was built to get the platoon schedule under the maximum fuel saving. The results show that: (a) The platoons' fuel- savings will increase with the increase of transportation missions and the maximum platoon length; the average driving distance and fuelsavings of vehicles in the platoons will eventually converge to a range; (b) The platoon traveled distance does not necessarily rise with the increase about maximum platoon length, i.e., the maximum fuel- saving strategy is not the maximum driving distance strategy for a spontaneous platoon; (c) More fuel-savings can be achieved by continuously increasing the spatial search scope without considering the temporal search scope; while increasing the temporal search scope cannot receive more fuel-savings when the spatial search scope is not considered; (d) 4000 trucks' missions could save up to 2026.21 L per day if they form platoons spontaneously; the average fuel-saving in the situation of platoons with five trucks can be improved by up to 13.92% compared with those of platoons with two trucks. Suppose this algorithm is applied to a larger-scale truck data set, the number of spontaneous platoons will increase significantly, and correspondingly, more fuel consumption and pollutant emissions can be reduced, which has excellent application potential