The Research On Optimization Of Slab Relocation Problem In Slab Warehouse

The slab warehouse is a transfer station to ensure the continuous production of continuous casting process and hot rolling process,which stores semi-finished slab that produced by continuous casting process and provides raw materials for hot rolling process.This thesis is on the background of slab stock logistics and extracts a new type of spatial distribution optimization problem of slab.Considering the characteristics and requirements of each logistics operation process such as slab storage,slab shuffling,discharging and stacking,the problem is to determine the execution order of different types of logistics operations and the spatial distribution of slabs.This study can improve the logistics equipment utilization and the regularity of the slab warehouse,reduce the amount of slab shuffling.Therefore,it is of great practical value to study the problem.Focus on the modeling and optimization of slab relocation problem in slab warehouse,the main work of this thesis is as follows:(1)Considering the constrains such as slab stacking,extraction and stacking regularity,aiming at minimizing the total number of shuffles and the overall crane traveling distance,a space-time network model is established.In this model,the moving phase of slab is taken as time axis and the storage position of slab is taken as the spatial axis,and the balanced constraints of network flow are established in linear style to describe the relations of temporal and spatial of the problem.The CPLEX of standard optimization software was used to solve the small scale problem of the model,which verifies the correctness and validity of the model.(2)Due to the complexity of the problem structure and model,CPLEX optimization software can only solve small-scale problems.In order to solve the actual problem of large-scale production,a differential evolution algorithm is designed to solve the problem.A heuristic algorithm based on the real-time status of different types of logistics operations is constructed to generate the initial population.Different from the general differential evolution algorithm,this paper proposes a separate encoding mechanism based on real number according to the structure of the problem.And a multi-population mutation strategy based on mutual learning among populations is designed to improve the result of this algorithm.The experimental results show that the proposed algorithm is effective.(3)Based on the mathematic model and the improved differential evolution algorithm,a decision support system about the optimization of slab relocation problem in slab warehouse is designed and developed.An integrated optimization method is provided in the system to guide the spatial allocation of the moving slabs in the various operations of the slab warehouse,which effectively reduces times of slab shuffling in the whole process of slab storage and the cost of business,improves the efficiency of slab warehouse operation.