A hybrid classical approach to a fixed-charged transportation problem

Abstract

Some of the most complex problems, nowadays, are transportation problems. A capacitated fixed-charge transportation problem is the problem we are trying to solve using hybrid classical approaches. The problem is considered with fixed capacities for each distribution centers and customers have particular demands. The model, as an economical model, minimizes the total cost as some distribution centers are selected in order to supply demands of all the customers. In order to find feasible solution for the mentioned problem, we are using some variants of Nearest Neighbor search algorithm. The problem as a whole is a two stages supply chain network: first we have to choose the distribution centers and next the customers based on their demand. The new approach is that we are starting from given customers' demands and select the best distribution centers. Some hybrid variants of Nearest Neighbor based on different probabilities are investigated and tested on large sizes data. Based on the numerical results we found a suitable hybrid version for the specified transportation problem. © 2012 Springer-Verlag.