Abstract: In order to satisfy the individual and various demands of customers, we establish a multi-model and multi-cargo loading problem model. Hybrid heuristic algorithm is used to get the optimization solution. Firstly,we use sequence of real numbers coding so as to simplify the problem. On the basis of cubage-weight balance algorithm, we construct an initial solution to improve the feasibility. Through adopting strategy combining sorting options with the best reserved, we ensure the diversity of population. We adopt the improved non-uniform mutation so as to enhance local search ability of chromosomes. Secondly, the elite population genetic algorithm seeks taboo search, to improve the search efficiency. Finally, an example is given to illustrate the effectiveness of the models and methods and it can offer an thought to settle the practical problem on a large scale.

Key words: operations research and cybernetics, loading problem of multi category goods, hybrid heuristic algorithm, stock elite, cubage-weight balance

摘要： 为满足客户多样化和个性化的需求,建立能充分、均衡利用装载工具的载重和容积的多品种、多车型货物配装模型,并从全局、整体最优上设计混合启发式算法求解。首先,采用实数序列编码,使问题变得更简洁;基于容重比平衡法构建初始解,提高了解的可行性;用基于排序的选择与最佳保留相结合策略,保证群体的多样性;采用改进的非一致变异,加强染色体的局部搜索能力;其次,对遗传算法求得的精英种群再进行禁忌搜索, 提高了搜索效率;最后,通过实例计算证明了上述模型和算法的有效性,并为大规模解决实际问题提供思路。

关键词: 运筹学与控制论, 零散货物配装, 混合启发式算法, 精英种群, 容重比平衡法