Research On Supply Chain Scheduling With Uncertain Processing Time In Cold Rolling

Cold rolling area in iron and steel enterprise is composed of multiple production stages.Each production stage includes multiple production lines,i.e.acid rolling,continuous annealing.The materials of the downstream production lines are supplied by their preceding production lines,thus the production lines in cold rolling area form a cold rolling supply chain.This thesis studies the supply chain scheduling with uncertain processing time in cold rolling from the integrated perspective of cold rolling supply chain,which is to determine the cold rolling schedule in order to reduce the production and inventory costs,considering the requirements of each production lines and uncertain parameters.This work resolves the conflicts among the requirements of each production lines by coordinating the schedules of each production lines,thus the product quality,the utilization efficiency of the production lines and the inventory structure are improved,and the inventory costs is reduced.Moreover,the uncertain parameters are concerned,thus more practical schedule can be achieved and the execution effect can be improved.The main contents of this thesis are given as follows:(1)A deterministic supply chain scheduling problem is derived from practical cold rolling supply chain,which is to determine the production sequence of the coils and their starting time,with consideration of each production line's requirements,the layout of the production lines and the cohesion relationship of different production lines.The problem is characterized by multiple production stages,multiple production lines.For this problem,a mixed integer linear programming model is established with the objective of minimizing the total changeover costs and penalties incurred by waiting time.Then the model is solved by CPLEX.This is the foundation work of solving the supply chain scheduling with uncertain processing time in cold rolling.(2)A supply chain scheduling problem with uncertain processing time in cold rolling is derived from the practical case with uncertain processing time due to various processing speed etc.To describe the uncertainty,a stochastic mathematical model is formulated by inducing stochastic parameters,with the objective of minimizing the total changeover costs and the expectation of penalties incurred by waiting time.Since the distribution of the stochastic parameters cannot be achieved,the expectation item in the stochastic model is approximated to average value by data sampling,based on the sample approximation method.(3)A sample average approximation method is proposed to solve the approximated model.The scenarios used in model are achieved by sampling.The improved differential evolution algorithm is developed to solve the model.In order to evaluate the solution,the lower bound of the original problem is calculated.Finally,the effectiveness of the algorithm is verified by experiments.(4)According to the practical requirement of cold rolling production,a supply chain scheduling system of cold rolling which has good human-computer interaction is designed and developed,based on the above model and algorithms.It can automatically generate cold-rolling production schedules to guide the production of cold-rolling plant.