Research On Integrated Optimization Of Production Scheduling For Automated Container Terminal

Automated container terminal(ACT)is a complex system with highly coordinated multiple subsystems in an unmanned environment.The resource optimization research involves many related fields such as management,operations research,information science,and port engineering.It's more sensitive to the coupling between subsystems than traditional container terminals,and has higher requirements for the effective management of core resources such as berths,quay cranes(QCs),AGVs,yard crane(YCs),and yard space.Its optimal scheduling requires consideration from the perspective of overall system integration.In this paper,the optimal scheduling of the ACT resources is taken as the research object.Guided by the integration idea and combining with the characteristics of its production operations,an optimization framework for the ACT production scheduling was proposed.Based on the theory of operations research and the technical means of computer simulation and intelligent algorithms,a multi-stage modeling method for integrated optimization of production scheduling of ACT was established.At the same time,according to the characteristics of the problem and the model,optimization algorithms for the global problem of container terminal production process were proposed.It further improves the theory and method for the optimized dispatching of ACTs,and also provides guidance for actual terminal managers.The research in this article has the following points:(1)Proposed an integrated optimization framework for production scheduling of ACTsBased on the system composition characteristics of the automated container terminal,the three aspects of its information physical composition,system core elements and system characteristics were comprehensively analyzed.Then,based on the integration idea and combining with the characteristics of its production operations,the integration theory of automated container terminal production scheduling was proposed,which includes the integration of production scheduling,the integration analysis method,and the integration optimization model.Finally,the scheduling problems in the berth subsystem,QC subsystem,AGV subsystem,yard management subsystem,and YC subsystem were analyzed and described in detail.The internal elements of each subsystem and the relationship between the subsystems were analyzed and discussed,including the acquisition,representation and evaluation of each efficiency characteristic element in the system.And an integrated optimization framework for the ACT production was established.(2)Established an integrated optimization model for shoreline resources of ACTThough the correlation analysis of ship and shore bridge allocation and the analysis of space-time conflicts in berth allocation,the ship handling time estimation formula mentioned in Meisel et al.^[1]and Elwany et al.^[2],which uses the maximum number of QCs to estimate,was analyzed.By adopting the strategy of replacing the maximum number of QCs by the allocable QCs,the rationality and accuracy of ship handling time estimation were improved.Then based on mathematical programming theory,an integrated optimization model for continuous berth and time-varying QC allocation was established with the goals of minimizing waiting penalty time,handling time,and QC conversion penalty time.Based on the spatio-temporal conflict characteristics of berth and QC allocation,a heuristic genetic algorithm based on spatial temporal conflict analysis was designed for the model.(3)Established a QC scheduling optimization model considering ship stability constraints,and completed the evaluation analysis of the QC interference indexThrough analyzing and describing the characteristics of the QC scheduling problem,a new QC scheduling model with minimizing the maximum completion time was established by integrating the the longitudinal stability constraints of ships into the the general QC scheduling model.A genetic algorithm based on heuristic was designed to solve the problem.Then,on the basis of this optimized model,a large-scale numerical simulation test was performed to evaluate and analyze the QC interference index in the ship handling time estimation formula(1-1).(4)Established an integrated optimization model for the yard resources of ACTThrough analyzing and describing the operating characteristics of the ACT yard and the flow characteristics of import and export containers in the yard,by transforming the optimization goal of the operation balance between blocks into constraints,an upper-level model of the yard block allocation with the shortest transportation distance was established to determine the allocation of yard blocks for all ships in the cycle and the allocation of different container types in the corresponding container yard blocks.Then,based on the upper model,a lower layer model for the distribution of bays with the goal of minimizing the moving distance of YC was established to solve the problem of the allocation of bays in the import and export blocks and the allocation of bays in the secondary transport.And a genetic algorithm based on the random value of the maximum and minimum intervals and a heuristic algorithm were designed to solve the upper and lower models respectively.Finally,by that the lower-level model heuristic algorithm was combined with the upper-level model solving algorithm in different ways,and embedded and tandem simulation processes were designed to solve the integrated optimization of the yard resources.(5)Established an integrated optimization model for the"QC-AGV-yard"in the ACT bi-directional operation modeBy analyzing and describing the characteristics of the import and export container flow between the ship and the yard,the task relationship of the container flow at each handling node and the time relationship of the corresponding tasks were established.Then,assuming that the storage location of the import and export container was known,an integrated scheduling optimization model of"QC-AGV-YC"with a minimum completion time was established.Based on this,the decision-making problem of the storage location of the import and export containers was further integrated into the"QC-AGV-YC"model,and the integrated optimization model of"QC-AGV-Yard"was established.Finally,combined with the time characteristics of the container at the loading and unloading node,two genetic algorithms based on heuristic,"shortest path"and"earliest acquisition time",were designed to solve the problem.The innovative research in this article is aimed at the shortcomings of the optimization scheduling in the ACT,and further establishes and improves the theory and method of the integrated dispatch optimization of the ACT.It provides research support for the optimized dispatch theory of automated container terminal systems,and also provides guidance for current or future automated container terminal managers.