oalogo2  

AUTHOR(S): 

Archil Prangishvili, Otar Shonia, Irakly Rodonaia, Artiom Merabian

 

TITLE

Adaptive Real-World Algorithm of Solving MDVRPTW (Multi Depots Vehicle Routing Planning with Time Windows) Problem

pdf PDF

KEYWORDS

vehicle routing planning, adaptive algorithms, congestion, multi-agent simulation, cloud computing, autonomic component, autonomic ensemble

ABSTRACT

The adaptive algorithm to solve MDVRPTW problem is proposed in the paper. Realistic real-world situations, such as presence of various congestion types on roads, are carefully considered and accounted for in the algorithm. To overcome the lack of realistic and reliable methods of congestion duration estimation we use the MatSim large-scale agent-based simulation tool. This tool allows users to compose and run complex simulation models that are extremely close to the real-world situations. Our approach implements also autonomic components ensembles concept. Each vehicle is associated with the corresponding autonomic component AC (a virtual machine in datacenter) and exchange on-line information with other vehicles. Besides, ACs can reschedule routes in order to find the acceptable alternative routes that enable vehicles to meet time windows requirements and, at the same time, avoid the congested roads. The adaptive algorithm is able to reschedule and find alternative routed for several vehicle in parallel, which increases the performance of proposed approach.

Cite this paper

Archil Prangishvili, Otar Shonia, Irakly Rodonaia, Artiom Merabian. (2017) Adaptive Real-World Algorithm of Solving MDVRPTW (Multi Depots Vehicle Routing Planning with Time Windows) Problem. International Journal of Transportation Systems, 2, 1-6