Modeling And Optimization For Enterprise-wide Production Scheduling In Rare Earth Process

Rare earth is a kind of crucial strategic resource for national economic development and security with wide applications in military defense,industrial production,petrochemical engineering,new materials,and so on.With a sharp decrease of the rare earth reserves,the contradiction between supply and demand is increasingly evident.Therefore,scientifically and efficiently improving the utilization rate of rare earth resources has become a focus of current research in order to settle the balance problem of supply and demand in rare earth production.This thesis focuses on the rare-earth production scheduling problem.Based on the process and technology of rare earth production process,a rare-earth production scheduling model is established based on unit-event continuous-time representation method.Considering the uncertainties of demand for rare earth products,the demand scenario tree of rare-earth product is introduced into the rare earth production scheduling problem with uncertain product demands.An improved Outer Approximation method with multi-generation cut is designed to solve the large-scale practical scheduling problem.Finally,a decision support system of rare-earth production scheduling optimization is developed for the rare earth production process.The detail work is as follows:1)Based on the properties of the whole process of rare earth production,the superstructure is adapted to describe the rare-earth whole production process.Making using of the continuous-time representation method,the MINLP model for rare-earth whole process production scheduling is formulated-Numerical experiments show the validity of this proposed model.2)According to the uncertainties of market demand for rare earth products,the mathematical programming is utilized to generate demand scenario trees for praseodymium oxide,neodymium oxide as well as praseodymium-neodymium oxides.We modal the scheduling model for rare earth production with the considered uncertainties of product demand.The maximum expected revenue over the scheduling horizon and the production scheduling solutions of different demand scenarios are obtained with solution the proposed model.3)Since the general Outer Approximation algorithm has a trouble in solving large-scale rare earth production scheduling problem,an improved multi-generation cut Outer Approximation algorithm is designed.The efficiency of Outer Approximation algorithm is improved with simultaneously solving multiple sub-problems and generating multiple cuts.Numerical experiments illustrate the effectiveness of the proposed improved algorithm.4)Based on the real-world demands of rare earth production scheduling management,an optimization decision support system for rare-earth production scheduling is designed and developed.The function modules include data management,parameter setting,scheduling scheme,human-computer interaction,report generation and system management.We can manage the data of materials as well as equipment,set production parameters,display production scheduling solution with Gantt chart,adjust scheduling solution and generate production reports.