现实约束下的多挂靠港滚装船舶配载优化

Optimization of Ro-Ro Ship Loading for Multiple Ports Based on Realistic Constraints

在分析滚装船舶配载特点的基础上,考虑了船舶稳性、配载效率以及航次收益等现实约束,针对多挂靠港的滚装汽车运输新模式下的配载优化问题进行建模,并基于该问题多阶段、多维约束、多背包组合优化的特点,开发了遗传算法对所构建模型进行求解.对于小规模数值算例实验,运用所设计遗传算法的求解结果与lingo软件所得到的精确解比较,平均误差率约为1.4%;对于大规模仿真算例,遗传算法的求解质量则明显优于现实调度策略,显示出本算法较好的鲁棒性.在此基础上开发了基于VB的配载信息系统,实现了配载过程及配载方案的可视化,为实时配载决策奠定了基础.

Based on the analysis of characteristics of ro-ro ship loading,the ro-ro ship loading problem for multiple ports is formulated as a mixed integer programming,considering realistic restrictions such as the stability of vessels,the efficiency of loading and the income of voyage.With the characteristics of multistage,multi-dimensional restrictions,multi-bag combinational optimization,a genetic algorithm （GA） is developed for solving the proposed model.As to the small-scale problems,the GA solved results are quite near to the exact solutions only by 1.4％ ; as to the large-scale problems,results from the GA are obviously better than those derived from practical dispatching rules,thus the robustness of the proposed algorithm is verified.Furthermore,a MIS based on Visual Basic is developed to realize the visualization of both loading process and loading patterns,which can provide some assistance for real-time decision-making.