Research On Multi-objective Hot Rolling Production Planning Problem Considering Material Flow Direction

Hot rolling is an important stage in iron and steel production.It is a transition stage from large-scale production to personalized production in steel production.It is closely related to the previous continuous casting process.At the same time,a part of the finished product in the hot rolling process is used as the raw material for the cold rolling process to ensure continuous production after hot rolling.Reasonable formulation of hot-rolled production plans can not only reduce the cost of hot-rolled production,but also effectively reduce the logistics costs of slab warehouses,which plays an important role in reducing costs and increasing efficiency of steel companies.Based on the actual hot rolling production planning process of a steel company,this thesis extracts the hot rolling production plan considering the material flow direction.Different from the existing hot rolling plan problem,this problem considers the material requirements of the cold rolling and the logistics efficiency of the slab warehouse,and formulates the rolling sequence of the hot rolling contract to meet the multiple objectives of production and material flow.This thesis studies the modeling and optimization methods of this problem.The main research contents are as follows:1)To solve this problem,a multi-objective integer programming model is established.The model considers the flow direction,process and logistics constraints such as material requirements,hot rolling process and slab storage stacking rules after hot rolling,with the goal of minimizing production switching costs,maximizing logistics efficiency and maximizing hot charging ratio.The slab in the unit and its rolling sequence.2)Aiming at the situation that the multi-objective optimization mathematical model optimization software proposed in this paper can’t directly solve the problem,a multi-objective discrete differential evolution algorithm is designed to solve the problem.In order to improve the performance of the algorithm,combined with the characteristics of the problem,the variation and selection of the algorithm are improved.First,three different mutation strategies are used in different search stages.Secondly,in the selection stage,new individuals are generated around elite individuals to enhance the search ability of the algorithm.Finally,based on the three metrics of multi-objective optimization problem,EP-MDE is compared with MDE,NSGA-Ⅱ and MOEA\D algorithms.Experiments show that the proposed algorithm has good performance on large-scale problems.3)According to the actual situation of hot rolling in iron and steel enterprises,a combination of random forest and support vector machine was proposed to determine the weight coefficient of each target in the hot rolling production planning problem,and the multi-objective problem was transformed into a single-objective problem.Through the training of the historical rolling data of the hot rolling plan,the weight coefficient of each target is optimized,the optimal solution is selected from the optimal solution set obtained by the multi-objective algorithm,and the actual hot rolling production plan is optimized.4)Based on the above proposed model and algorithm,the hot rolling production planning decision support system was designed and developed.The system can automatically prepare a hot rolling production plan and manually adjust it through a human-machine interface.The system interface is simple and easy to operate,which improves the efficiency of hot rolling planning.