基于混合智能算法的航空集装箱多目标装载优化

Multi-objective Optimization of Air Container Loading Based on Hybrid Intelligent Algorithm

目的针对航空集装箱多目标装载优化问题,综合提升航空货物的装载效率,降低装载成本,确保装载安全。方法结合航空货物装载实际情况,将容积利用率、载重利用率和重心偏移量作为优化目标,充分考虑载重、体积等7类约束,构建航空集装箱多目标装载优化模型。设计一种混合智能算法对该模型进行求解,先采用三空间分割策略结合改进遗传算法生成初步装载方案,再结合大规模多目标进化算法深度优化,综合考虑各目标函数值,生成最终的装载方案。结果采用国际BR算例和真实货物数据进行测试,实验结果表明,针对BR算例中不同异构型货物的平均填充率超过93%,在真实货物装载下,平均容积利用率和载重利用率分别为90.56%、76.40%,平均重心偏移量不超过5 cm。结论所提出的装载算法对航空集装箱多目标装载场景具有良好的适应性,有助于提升航空货物的运输效率和运输安全性。

The work aims to put focuses on optimizing air cargo loading efficiency,cost,and safety by finding optimal solutions for multi-objective container loading plans.A practical multi-objective model was established for aircraft container loading,focusing on volumetric efficiency,improving payload efficiency,and minimizing center of gravity shifts,while comprehensively considering seven types of constraints including weight limitations and volume restrictions.A hybrid intelligent algorithm was devised for the resolution of the given model.Initially,a tripartite spatial segmentation methodology was integrated with an enhanced Genetic Algorithm(GA)to produce an incipient loading configuration.Subsequently,the algorithm underwent extensive refinement through the application of a Large-scale Multi-objective Evolutionary Algorithm(LSMOEA),to ultimately yield the definitive loading scheme upon holistic consideration of the values associated with multiple objective functions.Validation was conducted with BR samples and authentic freight datasets.Results showed a mean packing density surpassing 93%for various cargo types in BR cases,with volumetric efficiency and payload utilization averaging at 90.56%and 76.40%in real shipments.Additionally,the mean center-of-gravity offset remained under 5 cm.The research affirms the algorithm's notable versatility in addressing multi-objective air container loading,significantly enhancing cargo transportation efficiency and safeguarding transit security.