The Model And Algorithm Of Slab Storage Problem In Slabyard

Slabyard is a product warehouse between steelmaking area and hot rolling area, it stores continuous casting slabs from the upstream steelmaking process, and supply slabs for the downstream hot rolling process. Operational efficiency of slabyard has an important impact on the efficiency of the overall production of steel plant.Slabyard operations mainly including slab storage, slab relocation and hot furnace scheduling. This paper studied the slab storage problem in slabyard (SSPS), an important problem in slabyard scheduling. In this problem, appropriate stack have to be selected for every inbound continuous casting slab to minimize relocations. With the minimization of relocations, the hot rolling production will become more efficiently.SSPS is a very complex combinatorial optimization problem, for the calculation of the objective of SSPS is another important combinatorial optimization problem in slabyard scheduling, the slab relocation problem. To reduce the complexity of the problem, this paper proposed two simple and effective methods to evaluate the objective of SSPS:the lower bound method and the reversion method. This paper also proposed three algorithms to solve SSPS:1) Mathematical programming algorithm based on 0-1 integer programming model. This paper first proposed a 0-1 integer programming model. After detailed analysis on the model, three optimization strategies and a relaxation method are proposed to simplify and speed up the model. A large number of experiments are designed to study the factor that affect the speed of the model, and the performance of the optimization strategies and relaxation method.2) Heuristic algorithm. Based on the idea of "Task-Resources-Income" model and the concept of descending subsequence, this paper proposed an innovative heuristic algorithm. A large number of experiments are designed to analysis the accuracy of the algorithm and its performance in large-scale problem solving.3) Rollout algorithm. Three simple basic heuristic methods are designed, and two Rollout algorithms are applied to improve their accuracy. A large number of experiments are designed to study the performance of Rollout algorithm, and compare the performance of different heuristic algorithms and different Rollout algorithms.