Abstract: Container loading sequence is a NP-complete problem.This paper constructs a mathematical model about shipping container in terminal，using Genetic Algorithms（GA） to test the model for determining the genetic algorithm parameters choice.Those factors about stability，reducing the number of container transfers and operability are decomposed into evaluating strategies.These strategies are classified according to PRI.In particular stability problem is decompounded into single row or cell for searching an optimal solution.Experimental results show that the algorithm can not only fulfill requirements of shipping but also optimize the loading sequence when reasonable parameters are chosen in the algorithm.

Key words: shipping sequence, stability, transfers, operability, algorithms design, GA parameter

摘要： 集装箱装船顺序问题是NP完全性问题，建立集装箱船配载数学模型，运用遗传算法，并实验确定适合该模型遗传算法的参数范围。将稳性、减少翻箱、可操作性等重要因素分解为评估策略，按优先级将各种评估策略划分等级，尤其将稳性问题具体到船的每行及每个载位，以寻找最优解。实验结果表明，只要将遗传算法中的参数选择合理，此方案则符合船舶配载要求，且集装箱装船顺序得以优化。

关键词: 装船顺序, 稳性, 翻箱, 可操作, 算法设计, 遗传算法参数