集装箱码头泊位计划干扰恢复多目标模型

Multi-objective Disruption Recovery Model for Berth Plan in Container Terminals

研究集装箱码头中干扰事件发生后泊位计划的调整问题,目的是降低干扰事件对集装箱码头作业系统的干扰。基于干扰管理方法,建立泊位计划干扰恢复多目标、多阶段模型,该模型考虑码头不同客户的特点以及多方利益的平衡,从码头作业成本、船舶延误以及计划偏离度三个方面度量系统扰动。为求解模型,提出了基于字典续的求解方法,并利用算例对模型与算法的有效性进行了验证。结果表明:该模型与算法可以有效解决泊位计划调整问题,模型能够考虑各方的利益以及码头各类客户的特点,因此得到的泊位调整方案更科学,同时,模型各目标的重要顺序可根据情况进行调整,实用性与可操作性更高。

The objective of the berth scheduling is to determine berthing times and positions of containerships in port container terminals. Accurate berth scheduling information can help minimize the penalty cost resulting from delays in the departures of vessels and additional handling costs resulted from non-optimal locations of vessels in a container terminal. Numerous studies have been conducted regarding efficiency improvement of berth operations in port container terminals. Many models and algorithms are developed to optimize the berth scheduling plan. However, during operation the planned schedule often has to be revised because of disruptions caused by severe weather, equipment failures., technical problems and other unforeseen events. Once these disruptions happen, the initial plan may become infeasible, and modification of current or future schedule should be undertaken to minimize the negative impact of the disruption. Such disruption management methods are needed to tackle any disruptions and obtain an optimal rescheduling scheme. In this paper, the problem of berth schedule recovery is studied to reduce the influence caused by disruptions. Based on the disruption management theory, we propose a multi-objective multi-stage model considering the characteristic of different customers and the tradeoff of all parties involved in the container terminal system. A new solution method incorporating the lexicographic optimization is proposed to deal with the multi-criteria nature of the problem. In the first part, the existing studies on container terminals and disruptions management are reviewed. Although numerous studies have been conducted, disruption recovery of berth plan is less studied. The berth plan recovery problem is NP-hard problem and complex constraints should be considered. Also, the characteristics and tradeoff of different parties should be considered. In the second part, the disruption management model for berth plan recovery is developed. The model is based on the method of multi-stage and multi-objective. The ships arrive at container terminals classified to three types, namely key line, general backbone line and branch line ships. A three-stage model is developed. The first stage is berth plan recovery model for key line ships. The second stage model is for general backbone line ships and the third stage is for branch line ships. The objective orders for different ship are different after considering the characteristics of each type of ships. In the third part, an algorithm based on lexicographic optimization and tabu search is designed. The algorithm includes three stages. For each stage, the level-by-level optimization method is used, and a hierarchical order among all the optimization objectives is established considering the importance of each objective. The optimal objective value of upper level is considered as a constraint of lower level, and tabu search is used to search neighbor solution. Finally, numerical experiments are provided to illustrate the validity of the proposed model and algorithms. Results indicate that the designed model and algorithm can efficiently tackle the berth plan recovery problem because that the tradeoff between all parties involved are considered. Our proposed model and algorithm are more flexible and feasible in all aspects of practical applications because decision makers can use them to adjust the objective order.