The Research And Application Of Balanced Loading Optimization Problem

Distributed loading is a key link in modern logistics. Reasonable distributed loading can save a lot of logistics costs, and improve the competitiveness of enterprises. So far, however, most of the domestic delivery centers is still distributing cargos to loading by past experience. In this way, they can't make better use of transport capacity of conveyances, which will increase the transportation costs. Moreover, with the increase of the load capacity of conveyances, in order to ensure transport security, the importance of balance load is increasing, which means the center of gravity of the goods on a conveyance should be located in a special range. Under this background, this thesis has deeply studied the Balanced Loading Optimization Problem (BLOP).First of all, BLOP is described and defined in this thesis. Based on the characters of different parts in the objective function of BLOP, a two-level algorithm is presented. In each level, a hybrid algorithm composed of a modern optimization method and a fast algorithm is developed. Applying different modern optimization methods to the two-level algorithm can get different combinations of the two level algorithmsSecondly, due to the conflict between the different parts in the objective function of BLOP, BLOP can be transformed into a multi-objective problem. Based on the traditional multi-objective genetic local search algorithm, a parallel multi-objective genetic local search algorithm is presented. The algorithm has successfully solved multi-objective BLOP.Finally, the steel coils loading problem of Baosteel is described and analyzed in detail. Baosteel has developed a number of special loading rules to strengthen transport security, it also reduces the difficulty of the problem. Therefore, for this special application case, a specific algorithm is designed, which is based on random selection, traverse and heuristic approach. Genetic-Simulated Annealing two-level algorithm and the specific algorithm are compared for solving different problems.