Research Of Three-dimensional Space Loading Problem Based On Cultural Algorithm

In the industrial field, we often solve problems about put some boxes into the container, loading is one of the important aspects especially in the logistics and transport industry, and it has a wide range of application. With China’s economic development, domestic and international freight is growing at a breakneck speed, logistics and transport is confronted with unprecedented challenges and opportunities. As an important step in logistics and transport process, to find a reasonable and effective way to solve the loading problem has a very important significance in reducing transportation cost, full and effective using space, guaranteeing the quality and efficiency of freight and improving the automation level of distribution business.First, this paper studies the three-dimensional loading problem. According to the different forms of problem, we classified into three types to describe three-dimensional loading problem, introduced the different three-dimensional structure partitioning strategy and methods of solving problem when loading. By analyzing the advantages and disadvantages of various solving algorithms, we point out that using intelligent optimization algorithms more quickly and effectively when deal with large-scale loading problem, and stability is also very good.Loading problem is a typical NP-hard combinatorial optimization problem, with genetic algorithms as a symbol, intelligent algorithms are widely used for solving constrained optimization problems currently, but these algorithms are only based on population space search, its efficiency is not high, and fall into local optimum. For the characteristics of research in this paper and the limitations of traditional algorithms, we proposed cultural algorithm which based on double evolutionary mechanisms influence and promote each other to solve the problem. Afterward, described in detail the space structure of cultural algorithm, and the communication protocol between double layer structures.Analysis shows that the experience in belief space of cultural algorithm will be update after influence a generation population space, there is no difference between different belief space, the outstanding experience can’t be saved and will be replaced by next generation, we propose belief sub-space cultural algorithm solving nonlinear constrained optimization problems, the belief space is divided into multiple sub-space, each sub-space saving outstanding individuals selected from one populations generation, and each group of outstanding individuals have different advantage, update sub-space according to priority and compare results when latest generation outstanding individuals selected from population space, the comparison conduce outstanding knowledge to be saved, and different advantage individuals not only promote the search for the optimal solution, but also keep population diversity.At the end of this chapter, we test and verify the performance of algorithm by function simulation experiments.Finally, set up a mathematical model combined with constraints in actual loading process, experimental and research multilayer cultural algorithm solving three-dimensional loading problem on MATLAB. The analyze of multiple groups of data simulation results and the comparison with traditional algorithm show that this algorithm is well used in solving three-dimensional loading problem, has some reference value to loading part in logistics and transport.