Berth Allocation Problem With Approach Channel And Tidal Constraints

Terminals,as hubs of transportation,play an important role in promoting economic development.With the rapid progress of information technology,the automation and intelligence of terminals have become a trend.This thesis studies the berth allocation problem in terminal intelligence.This problem is a variety of the two-dimensional bin packing problem in the combinatorial optimization.In the first chapter of this thesis,we introduces several important optimization problems in the terminal according to the port space layout,and reviews the three types of discrete,continuous,and mixed berth allocation problems based on the shoreline space factors.In the second chapter of this thesis,considering the factors such as cargo properties and vessel types,the continuous berth allocation problem with one-way channel constraints is studied,and a mixed integer linear programming model is established.The model can deal with berth allocation problems with multiple terminals,multiple vessels and channel constraints.We performed programming experiments using the optimizer Gurobi.The experimental results show that the model can be solved quickly when the number of vessels does not exceed 50,and the results are in line with the actual operation requirements.Considering that the vessel’s operating time is directly related to the number of quay cranes,in the third chapter of this thesis,we links the berth and quay crane allocation.We considered the joint berth-crane allocation problem which is limited by the tide factor and the influence of the channel,and divided the vessels into two types:tide and free.We simulated the high and low tides according to the actual situation,and established a mixed integer linear programming model.When the berth is allocated,the model dynamically adjusts the number of quay cranes of each vessel according to the remaining workload of the vessel and its own physical limitation,taking into account the balance and operating efficiency of the vessel.In solving the model,we use the optimizer Gurobi to solve small-scale examples.For large-scale examples,we introduce genetic algorithms,simulated annealing algorithms,and particle swarm algorithms to solve problems,and design the key steps in the solution process.Finally,a programming experiment is performed in this thesis.The experimental results show that compared with Gurobi,the proposed algorithm can save at least about 95% of the calculation time in small-scale examples,and the average calculation deviation does not exceed 2.45%.In the large-scale examples,the algorithm can handle calculation scales that the optimizers cannot process,and the average deviation obtained by the three algorithms does not exceed 10 %.