基于发达度的远程自主代客泊车乘车点布局优化方法
首发时间:2021-05-28
摘要:本文提出了一个两阶段的远程自动代客泊车(LAVP)系统下的乘车点布局优化方法。该方法首先评估备选乘车点的发达度,进而确定备选乘车点集合。综合考虑远程自主代客泊车系统的用户和运营商的双方利益,提出了一个以无人驾驶汽车平均等待时间最短、用户平均行程花费时间最小、无人驾驶汽车平均总行驶时间最短的多目标乘车点布局优化模型,以更好地服务于城市的远程自主代客泊车系统。最后以赫尔辛基主城区为例进行仿真实验,对本文提出的模型和算法进行检验,通过仿真实验结果分析发现,本文提出的乘车点筛选模型得到的乘车点使远程自主代客泊车系统的全部无人驾驶汽车无需等待停车位,验证了以发达度评价备选乘车点的方法是有效的,可以有效地为远程自主代客泊车乘车点布局优化提供合理的搜索空间。同时,从乘车点布局优化结果来看,本文提出的多目标优化模型可以兼顾乘车点的数量和位置及建设成本,优化结果符合赫尔辛基城市的实际情况。
For information in English, please click here
Pick-up Spot Optimization of LAVP Based on Connectivity
Abstract:In this paper, a two-stage method of drop-off and pick-up (D/P) spots optimization in long-range autonomous valet parking (LAVP) system is proposed. Firstly, the alternative D/P spots are determined by evaluating the development of each D/P spot. Based on the alternative D/P spots set, considering the benefit of users and the operator, a multi-objective model is presented to deploy the D/P spots for serving the LAVP system. The objectives include minimizing average parking waiting time, minimizing average automatic vehicle (AV) traveling time and minimizing average user trip duration. Finally, the two-stage method for D/P spots optimization is validated by simulation in Helsinki city. The results show that all AVs do not need to wait for parking spaces, which is verified that the method of evaluating the alternative D/Ps based on development is effective, which can effectively provide a reasonable search space for D/Ps. Furthermore, the solution of the D/Ps optimization accords with the actual conditions of Helsinki. This indicates that the proposed multi-objective model can consider the quantity, location and the cost of D/Ps\' construction.
Keywords: autonomous valet parking (AVP) autonomousvehicle (AV) location optimization attractiveness
基金:
引用
No.****
同行评议
共计0人参与
勘误表
基于发达度的远程自主代客泊车乘车点布局优化方法
评论
全部评论