Research On Batching Planning In Supply Chain Of Steel Industry

With the advance of the integration management of the steel industry, the core of the production management is changing from the optimization planning for single unit of one process to the integrated optimization planning for multiple units in plant-wide process. The production planning of the steelmaking stage is the essential part for make-to-order philosophy in an integrated iron and steel industry. Moreover, the production facilities are often use for batching production in the steelmaking stage. To increase the production operations management level, it has both practical and theoretical significance to study the steelmaking batching problem from the perspectives of the supply chain where the plant-wide process production constraint conditions are considered. The main contents of this thesis are as follows:(1) The new features arising in the supply chain batching planning problem are analyzed, including:(a) multiple units:which means the equilibrium between multiple steel plants and multiple casters, and the process routes optimization between the hot rolling and the steelmaking are considered; (b) multiple stage which means the impact of batching plan of the steelmaking stage on the production planning of the hot rolling and the cold rolling stage are considered; (c) multiple planning period which means the impact of batching plan of the steelmaking stage on the customer demands and the inventory fluctuations in the longer planning horizon are considered. Aiming at the above new features and the batch processing features of steelmaking and continuous casting process, we take as production quantity of each order to each caster on each day, the grouping relation from orders to charges and the grouping relation from charges to casters as decision variables and formulate a mixed integer programming model for the problem. The objective consists of the extra production cost due to batching operations and the penalty cost due to unbalance production on the downstream units of steelmaking and continuous-casting.(2) Based on the detail analysis of the features and complexities of the mathematical model, this develops the differential evolutionary algorithm to solve the problem. The components of the algorithm including encoding, decoding, crossover, mutation and adaptation value calculation are all well designed and improved according to the problem features. The computational experiments are carried out on practical data collected from steel company to test the performance of our algorithm. The results verify the practicality and the efficiency of our algorithm as compared with the manual planning method.(3) A decision support system for the supply chain batching plan in steel industry is designed and developed by embedding the above proposed model and algorithm as core solver. The system consists of data management module, model solver module, result editor interface module, and etc. Furthermore, we also develop a decision support system of shuffling optimization in the hot rolling process with the previous theoretical result support of our research team which consists of parameters configuration module, graphical interface of shuffling instruction, result analysis module, and etc. Both systems have achieved the preliminary application verification based on the particle data of a large domestic steel industry.