Research On Vessel Unloading Scheduling Problem In Bulk Cargo Port Based On Branch And Price Algorithm

The bulk cargo port and stockyard of iron-steel plant is one of main logistics nodes for the input and storage process of bulk cargo such as iron ore and coal,which is restricted by a series of operational resource such as capacity of berth and belt conveyor.As one of the main bottlenecks of the raw material logistics process,the operational efficiency of the bulk cargo port affects not only the logistics cost but also the subsequent production which can be carried out safely and continuously in the future.Formulating reasonable and efficient vessel unloading scheduling schemes can significantly improve the operation efficiency of input equipment and stockyard as well as reducing logistics costs,which has an important practical significance and economic value for the raw material logistics of iron-steel plant.The vessel unloading schedule of the stockyard is based on information of vessels,such as the arrival time,the weight of cargos and the occupancy of input equipment(berth,belt conveyor,vessel unloader,etc.),the stockyard layout and other operational resource information,for example,emergency demand of raw material.The above information is used to determine the unloading time of the incoming vessels and to allocate the operating equipments and loaction of stockyard.This thesis separately studies the process of raw material input in two different configurations(rive transportation and sea transportation)of a large domestic iron-steel plant,according to the actual situation and different demand of each stockyard to establish the belt conveyor allocation model and the model of vessel unloading problem with stock row allocation.Then use column generation algorithm to solve the linear relaxation problem and combin column generation and branch and bound to build the branch and price algorithm.The main research contents of this thesis are as follows:1)As for the vessel unloading scheduling problem with a space-limited stockyard for an iron-steel plant,and because of the storage capacity of the stockyard and the limitation of the depth of water by the Changjang River,it leads to a large number of small batch inputs,the storage capacity constraint and raw material balance of the stockyard are considered.To minimize the vessels’ demurrage cost as the objective function,a mixed integer linear programming(MILP)model is established.The problem of the original model is transformed into a set partitioning model and a series of price subproblem models by using the Danzig-Wolfe decomposition.In the iteration process of the column generation algorithm,a heuristic algorithm based on the first-come first-served idea is designed to obtain the initial feasible solution.By analyzing the characteristics of the price subproblem,an improved enumeration strategy is proposed to generate the minimal negative solution,thus improving the efficiency of colum generation algorithm.Base on the optimal solution of column generation algorithm,the non-integer optimal solution of the linear relaxation problem is needed to be branched.So the branch and price algorithm is constructed to obtain the optimal integer solution of the problem.Comparing the experimental results obtained by using the branch and price algorithm proposed in this thesis and the CLPEX solving software,the thesis wants to verify the correctness and effectiveness of the branch and price algorithm.2)Considering the vessel unloading scheduling problem with stock row allocation,and according to the long unloading time of the vessel and the use of one vessel multi-material transportation,the configuration of combination stacker reclaimer causing serious conflicts between the stacking and reclaiming operations,a mixed integer linear programming model is established to minimize the logistics cost incurred during the raw material input process(vessel departure cost,delay fine,and the cost caused by switching the belt conveyor route).The original model was reconstructed by using the Danzig-Wolfe decomposition and solved by designing a column generation algorithm.In order to get the optimal integer solution,a branching strategy based on the decision variable of the original model is designed.Then design an improved branching strategy based on two variables branch alternately is put forward.Then compare the effect of two branching strategies through data experiments.The correctness and effectiveness of the branch and price algorithm is verified by comparing the algorithm with the experimental results obtained by the CLPEX solving software.3)Develop the decision support system with the above model and algorithm as the core.Design and develop the data analysis and logistics dispatch optimization system for stockyard which can provide scientific and quantitative decision support for the raw material input process.This system can help iron-steel plants to improve the turnover rate of the raw material input system,avoid stacking and reclaiming conflicts and reduce the cost of logistics and other effects.