论文部分内容阅读
针对罐式集装箱运输的特殊性,考虑重箱流和空箱流调配的综合优化,以罐箱运输费用最小为目标,建立铁路罐式集装箱空箱调配优化多商品网络流模型,并构造了1种嵌入模拟退火操作的遗传算法对之进行求解。为了使模型与算法可得到更符合实际、操作性更强的结果,给出了3种空罐箱调配的策略,作为隐含条件加入到算法求解过程中。利用自适应遗传模拟退火算法对随机生成的实际规模问题算例进行求解,并与用通用代数建模系统软件GAMS的计算结果进行对比。结果表明,前者得出的结果与最优解差距不大,而且运算速度更快,更能满足解决实际问题的需要,为铁路罐箱调配优化提供了良好的决策支持模型和算法。
Aiming at the particularity of tank container transportation and considering the optimization of heavy tank flow and empty tank flow deployment, a multi-commodity network flow model of tank tank empty tank deployment was established with the goal of minimizing tank transportation cost. Genetic algorithms embedded in simulated annealing operations are solved. In order to make the model and the algorithm obtain more practical and operational results, three kinds of strategies for empty tank allocation are given, which are added as implicit conditions to the algorithm solving process. An adaptive genetic simulated annealing algorithm was used to solve the randomly generated real scale problem and compared with the results of GAMS. The results show that the former results are not much different from the optimal solution, and the computation speed is faster, which can meet the need of solving practical problems and provide a good decision support model and algorithm for railway tank allocation.