Research On Modeling And Optimization For The Logistics Supply Chain Scheduling In Container Terminals

Modern logistics industry has become an important industry in our country, and it has developed rapidly. Container terminal is a typical logistics company. The size of containership and the container throughput are getting larger, which make it difficult for container terminals because of the limited space and logistics equipments, so it is urgent for container terminals to schedule the space and logistics equipments effectively, in order to improve the productivity and the delicacy management level. Supply chain management is a hot research field in the past ten years, and its research mostly focused on supply, production and delivery, for the research on logistics supply chain scheduling in container terminals is less. There is tight connection between different stages in logistics supply chain. The independent scheduling at different stages often leads to low operation efficiency and high operating costs. We study the modeling and optimization for several logistics supply chain scheduling problems in container terminals. The effective scheduling in logistic supply chain will improve the efficiency of logistics operation, reduce the waiting time between various logistics equipments as well as the logistics costs, and improve service level of container terminals and enterprise competitiveness.Crane is a kind of important logistics equipment used to pick up, delivery, storage and retrieve containers in container terminals, and the crane scheduling is one of the bottlenecks that restrict the development of container terminals. We studied several logistics supply chain scheduling problems related to crane (quay crane, yard crane), which are the berth and quay crane supply chain scheduling problem; the quay crane and truck supply chain scheduling problem with unidirectional flow; the quay crane and truck supply chain scheduling problem with bidirectional flow; and the reshuffle and yard crane supply chain scheduling problem. We formulate the models and propose some algorithms for the above problems. The major researches are as follows:1) The berth and quay crane supply chain scheduling problem in container terminals is studied. The problem is defined as follows:under the condition that the length of each berth, the number of available quay cranes in each berth, and the arriving time of each ship are all known in advance, we decide the berthing time, berthing position and the number of quay crane allocated for each ship. Different from the existing studies, the berths are not in the same line, and appear in the shape of "F", "L", which contains several straight lines. As a result, some adjacent berths can’t be occupied by the same ship, and quay cranes cannot move along these berths. The model is formulated as a mixed integer programming model, and a lagrangian relaxation algorithm is proposed to solve the problem. The computation time is decreased by proposing two properties of the optimal solutions and a speeding up strategy. The experiment results show that the dual gap of the proposed lagrangian relaxation algorithm is 1.74% and 1.799% for the small and large sized problems respectively.2) The quay crane and truck supply chain scheduling problem only contained inbound containers in container terminals is studied. The problem is defined as follows: under the condition that the unloading time and transporting time of each container are known in advance, we decide the quay crane allocation and truck allocation, as well as the handling sequence on each quay crane and each truck. Since there are only inbound containers, the container transporting route only exists from the quay space to the yard, which is a unidirectional flow problem. The non-crossing constraints between different quay cranes and the precedence between different containers are considered. The objective is to minimize the completion time for unloading all the containers. A property of the optimal solutions, several valid inequalities and two lower bounds are proposed by observing the problem. A mixed integer linear programming model is formulated for the problem, and a particle swarm optimization (PSO) algorithm based on a new velocity updating strategy is developed for solving the problem. The research of the above problem is extended, and a model of the quay crane, truck and yard crane three levels supply chain scheduling problem is formulated, besides, the algorithm for solving the problem is designed. When the number of the one equipment among quay crane, truck and yard crane is fixed, and numbers of the other two equipments are different, the average deviation of the proposed algorithm and the optimal solutions are 1.676%,2.033% and 0.338%, respectively.3) The quay crane and truck supply chain scheduling problem in container terminals considering both inbound and outbound containers is studied. During the loading and unloading process for container ships, there often both exists inbound containers and outbound containers. In this case, truck can return quay space without any containers after transporting a container to the yard, and it can return quay space with a loaded container from the yard. During the whole scheduling process, there are two truck transporting routes. The problem is a bidirectional flow problem. The model, lower bound and algorithm of the unidirectional flow problem are modified for the bidirectional flow problem with both inbound and outbound containers. The computational results show that the average deviation of the proposed PSO algorithm and the optimal solution is 1.26%, and the average deviation of the proposed PSO algorithm and the lower bound is 2.48%.4) The reshuffle and yard crane supply chain scheduling problem in container terminals is studied. The problem is defined as follows:under the condition that the configuration of a bay and the time for external trucks to retrieve containers in the bay are known in advance, we decide the sequence for the yard crane to pick up all the containers in the bay and the objective position of each reshuffled container simultaneously. A mixed integrated linear programming model is formulated, and the complicated relationship between the number of reshuffle and the retrieving sequence of containers is described by linear expression. An improved PSO algorithm is proposed for solving the problem used a strategy reinitializing some particles. Compared with the results of the reshuffle scheduling problem and the yard crane scheduling problem separately, the computational results show that the average deviation of the reshuffle and yard crane supply chain scheduling and the separate scheduling problem is 1.84%.