Synchronizing vehicles for multi-vehicle and one-cargo transportation. (May 2018)

• Record Type: Journal Article
• Authors: Hu, Zhi-Hua; Wei, Chen
• Is Part Of: Computers & industrial engineering. Volume 119(2018)
• Page Start: 36
• Page End: 49
• Publication Date: 2018-05
• Subjects: Logistics management -- Vehicle routing problem -- Big-size cargo transportation -- Synchronization -- Genetic algorithm; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2018.03.023
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: Synchronize multiple unit-load vehicles for transporting cargos of various sizes. A sequential insert algorithm is developed based on practical experiences. A fast greedy insert algorithm is developed for large-scale instances. A genetic algorithm is developed based on the sequential insert algorithm. The synchronization parameters and algorithm performances are analyzed. Abstract: In ship manufacturing and big-size cargo transportation, several flat vehicles are usually cooperated to transport one big-size ship segment or cargo, namely "multi-vehicle and one-cargo transportation" (MVOC). It is distinctly different from general transportation scenarios where a vehicle is used to load and transport several cargos, namely "one-vehicle and multi-cargo transportation" (OVMC). The MVOC generates the difficulty in synchronizing several vehicles for transporting a cargo. A mathematical program that can be solved by existing mixed-integer linear program solver is formulated, by considering the synchronization constraint among the unit-load flat vehicles under the minimization of makespan. To improve the computation performance of the solution methods, a sequential insert algorithm is developed as the basic procedure for a greedy insert algorithm and as the decoding scheme of a proposed genetic algorithm. These three methods (the mathematical program, the greedy insert algorithm and the genetic algorithm) are compared by numerical studies considering the effects of instance scales and synchronization complexity on the optimality and computation performances. The genetic algorithm performed better for solving small- and medium-scale instances with lower complexity, and the greedy insert algorithm is very fast and suitable for solving large-scale and complex instances.