Clustered tabu search optimization for reservation-based shared autonomous vehicles
Clustered tabu search optimization for reservation-based shared autonomous vehicles
复制标题
DOI：
10.1080/19427867.2020.1824309
复制
发表时间：
2020-09
期刊：
影响因子：
--
通讯作者：
Shunhu Su;Emmanouil Chaniotakis;Santhanakrishnan Narayanan;Hai Jiang;C. Antoniou
中科院分区：
文献类型：
--
作者：
Shunhu Su;Emmanouil Chaniotakis;Santhanakrishnan Narayanan;Hai Jiang;C. Antoniou
摘要研究了基于预约的自动汽车共享系统的优化问题，目标是最小化车辆总行驶时间和顾客等待时间。因此，RACS系统和它的路由制定与系统效率和乘客的关注的考虑。一个元启发式禁忌搜索方法的研究作为一种解决方案的方法，结合K-均值（KMN禁忌）或K-中心（KMD禁忌）聚类算法。所提出的解决方案的算法进行了测试，在两个不同的网络不同的复杂性，并评估算法的性能。评价结果表明，TS方法更适合于小规模问题，而KMD-Tabu适合于大规模问题。然而，KMN-Tabu具有最少的计算时间，虽然解决方案的质量较低。
ABSTRACT This paper investigates the optimization of Reservation-based Autonomous Car Sharing (RACS) systems, aiming at minimizing the total vehicle travel time and customer waiting time. Thus, the RACS system and its routing are formulated with a consideration for system efficiency and passengers’ concerns. A meta-heuristic Tabu search method is investigated as a solution approach, in combination with K–Means (KMN–Tabu) or K–Medoids (KMD–Tabu) clustering algorithms. The proposed solution algorithms are tested in two different networks of varying complexity, and the performance of the algorithms is evaluated. The evaluation results show that the TS method is more suitable for small-scale problems, while KMD–Tabu is suitable for large-scale problems. However, KMN-Tabu has the least computation time, although the solution quality is lower.