不确定因素下的集装箱码头车船间装卸作业集成调度  

Integrated scheduling of handling operation between train and vessel in container terminal under uncertain factor

作　　者：常祎妹[1] 朱晓宁[1] 

机构地区：[1]北京交通大学交通运输学院,北京100044

出　　处：《交通运输工程学报》2017年第6期115-124,共10页Journal of Traffic and Transportation Engineering

基　　金：国家自然科学基金项目(71390332);中央高校基本科研业务费专项资金项目(2017YJS100);民航科技项目(201501)

摘　　要：为了提高集装箱码头的效率,建立了集装箱码头车船装卸作业集成调度模型;考虑了装卸同步作业与一些实际约束条件,如岸桥、龙门吊间干扰和安全距离等,以及生产调度中的不确定因素,如岸桥、集卡的速度变化等,设计了改进的多层遗传算法求解模型;为了确定随机变量的具体值,在传统的遗传算法中引入蒙特卡罗法,为了使得集装箱在集卡间均衡分配,在传统的遗传算法中引入均衡算子;通过试验验证模型和算法的可行性和有效性,并进行了改进的多层遗传算法中引入均衡算子与未引入均衡算子间的对比试验。分析结果表明:引入均衡算子后的装卸作业完成总时间为7 437.8s,比未引入均衡算子减少760.3s,且在引入均衡算子的试验中50FEU集装箱均匀分配给了5辆集卡,而在未引入均衡算子的试验中未能均匀分配;固定龙门吊和岸桥数量,改变集卡的数量,通过比较每组试验中作业总完成时间和边际效益值,得出当集装箱数量为80、90、100FEU时,最优设备配比为3台龙门吊、10辆集卡和4台岸桥。可见,改进的多层遗传算法可以提高集装箱码头车船装卸作业的效率,是解决集装箱装卸作业集成调度问题的一种有效方法。To improve handling operations the efficiency of container terminals, an integrated scheduling model of between trains and vessels was established. Loading and unloading synchronization, as well as some practical constraints, such as quay crane （QC） interference, gantry crane （OC） interference, QC safety distance, GC safety distance and so on, were considered. Some uncertain factors during production scheduling, such as the change of QC speed and inner truck （IT） speed and so on, were also considered. In order to solve the model, an improved multi-layer genetic algorithm （MLGA） was developed. To determine the value of random variable, Monte Carlo method was introduced into the traditional genetic algorithm. To balance the allocation of containers among ITs, a balanced operator was introduced into traditional genetic algorithm. The feasibility and effectiveness of the proposed model and algorithm were verified by using computational experiments, and a comparison experiment wascarried out between the improved MLGAs with and without balanced operator. Analysis result indicates that the makespan of handling operation with balanced operator is 7 437.8 s and 760.3 s less than that without balanced operator. Meanwhile 50 FEU containers are evenly allocated to 5 ITs in the experiment with balanced operator, while in the experiment without balanced operator, they are unevenly assigned to ITs. Fixing the number of GCs and QCs, and changing the number of ITs, when the number of containers is 80, 90, and 100 FEU, respectively, the best equipment proportions are 3 GCs, 10 ITs and 4 QCs by comparing the makespans and marginal benefit values in each group of experiments. The efficiency of handling operations between trains and vessels in container terminals is enhanced by using improved MLGA, and the improved MLGA can be an effective approach to solve the integrated scheduling problem of container handling operation. 6 tabs, 9 figs, 25 refs.

关 键 词：综合运输 铁水联运 集装箱装卸作业 集成调度 改进的多层遗传算法 铁路作业区 

分 类 号：U169.6[交通运输工程]