Robust Zinc Concentrate Procurement Method And System Implementation With Uncertain Raw Material Supply And Production Plan

The procurement of zinc concentrate is the core part of zinc smelters,providing the necessary raw materials for production,which directly affects the production and operation of enterprises.In terms of production,under the continuous production mode of zinc smelting,the recovery cost of shortage is huge,and the quality of zinc concentrate directly affects production efficiency and product quality.The production process has extremely high requirements on the supply stability and quality of zinc concentrate.In terms of operation,the procurement cost of zinc concentrate accounts for about 75% of the total cost,with a large capital occupation.Reducing procurement costs is very important for improving profits.However,due to the uncertain changes in zinc concentrate price and production plan,the procurement cost of zinc concentrate is high and there is a high risk of shortage.Furthermore,the supply quality of zinc concentrate varies greatly,and there are uncertain fluctuations in the arrival grade,making it difficult for the comprehensive quality of zinc concentrate to meet the production requirements.Robust and efficient zinc concentrate procurement decision-making is an important means to improve production stability and operational economy.Based on deep analysis of the changes and impact characteristics of various uncertainties,this thesis systematically studies the robust zinc concentrate procurement method from three aspects: raw material price forecasting,zinc concentrate total procurement amount decisionmaking,and supplier order allocation.On the premise of ensuring the continuous and stable supply of zinc concentrate and the comprehensive quality of the arrival,the procurement cost can be reduced,which is of great significance to the high-quality development of enterprises.The main research work and the innovations of this thesis are as follows:(1)A zinc concentrate price interval forecasting method based on multifactor selection and fusion is proposedIn view of the problems that the zinc product price is greatly affected by external factors,the causal relationship is complex,and the forecasting reliability is poor,the influence of external factors such as economic situation,supply and demand relationship,and industry development trend on price is analyzed.A method based on Granger causality test and XGBoost for causal relationship and importance analysis of external factors is proposed to screen out key external factors that affect price changes.Considering the poor training performance of forecasting model under limited price samples,a multi-factor fusion multi-step zinc product price forecasting method is proposed based on convolutional neural network and gated recurrent unit network.The confidence interval for future price is constructed based on Bootstrap framework and t distribution.Based on zinc product price and smelting processing fee,according to the price sharing mechanism and interval arithmetic law,the future change range of zinc concentrate price is calculated,providing information support for zinc concentrate procurement decision-making.(2)A robust zinc concentrate procurement decision-making method with uncertain raw material price and production plan is proposedAiming at the high procurement cost and high risk of shortage caused by uncertain fluctuations in zinc concentrate price and production plan,the inventory update model for zinc concentrate is analyzed,and a multi-period robust procurement model for zinc concentrate with uncertain raw material price and production plan is proposed.Meanwhile,the dynamic fluctuations in raw material price and the seasonal changes in production plan is analyzed,and an adaptive construction method for budget-based uncertainty sets is proposed to achieve accurate and adaptive description of uncertainties.Furthermore,for the problem that the robust procurement model is difficult to solve for semi-infinite programming,a robust counterpart that can be quickly and accurately solved based on duality theory is derived,and the total procurement amount of zinc concentrate that takes into account production reliability and operational economy under internal and external uncertainty is obtained.The procurement cost can be reduced on the premise of ensuring a stable supply of raw materials.(3)A supplier order allocation method based on multi-objective chance constrained programming is proposedDue to the large number of zinc concentrate suppliers and significant differences in supply characteristics,it is difficult to conduct comprehensive evaluation.A multi-level evaluation index system for zinc concentrate suppliers is established,and a method for determining the importance of evaluation indicators by combining data knowledge and expert experience knowledge is proposed.A weight coefficient selection method based on compatibility and difference degree is proposed,and the comprehensive performance and corresponding ranking of each supplier are automatically obtained based on TOPSIS.Furthermore,in response to the large deviation in the arrival grade of zinc concentrate,which makes it difficult to ensure comprehensive quality,the distribution of arrival grade deviation for different suppliers and elements is identified based on Gaussian mixture model.Taking into account for factors such as the knowledge of ore blending and supplier performance,a multi-objective chance constrained order allocation model is proposed that minimizes order cost and maximizes supplier utility.On this basis,an intelligent hybrid optimization algorithm combining Monte Carlo simulation,multi-objective particle swarm optimization and TOPSIS is proposed to solve the model and obtain robust ordering decisions under complex resource constraints.(4)An integrated optimization system for zinc smelting raw material procurement-production is designed and developedThe actual requirements of the zinc smelting raw material procurementproduction integrated optimization system are analyzed,and the system architecture based on Industrial Internet of Things is designed.The decoupling between the algorithm and the software is realized,which greatly improves the efficiency of software development and maintenance.The zinc smelting raw material procurement-production integrated optimization system is developed,which provides strong support for efficient procurement decision-making.It is of great significance for improving the supply chain decision-making level and enhancing the competitiveness of enterprises.