Research On Cost Optimization Of Lean Iron Ore Dressing Process

Lean iron ore mining is of great value to the development of the steel industry and the security of strategic resources in our country.Lean iron ore dressing is a complex system engineering.It is difficult to build a production process model for iron ore dressing due to the uncertainty of ore material,the multi-modal properties of production process and operating condition,and the variability of organization and management mode in production process.The production process of lean iron dressing presents systematic and dynamic evolution with multi-factor changing,interlaced discrete and continuous processes,and multi-system coupling.The management and control of production process in lean iron ore dressing show spatio-temporal non-linear characteristics with multi-level progress,multi-system coupling,and multi-objective coevolution.The lean iron ore dressing production cost system is crucial for ore dressing system.The existing cost accounting control method and technology for lean iron ore dressing are difficult to meet the demand of digital and intelligent development.The cost of lean iron ore dressing is lack of corresponding theoretiacal methods and technologies;is difficult to obtain multi-dimensional,dynamic,and elaborate cost data in ore dressing;is difficult to build the structure model of ore dressing cost system;is difficult to form a optimization method combined with priduction.How to optimize the cost system of lean iron ore dressing is not only on the key issue of cost management and optimization in ore dressing but also on the basic and common technical challenge in ore dressing engineering.It is critical to practical production in lean iron ore dressing.This research is based on the coupling relationship between lean ore dressing cost and ore dressing production process and faced the optimization objective of cost system in lean iron ore dressing production.Based on the cost optimization mechanism in ore dressing production,the cost optimization method for selecting ore raw materials and the data-driven cost optimization model for ore dressing production are studied using combinatorial optimization and deep learning.It is a new cost optimization method setting in the digital mine.This research mainly completed the following works:1)Analyzing the research status of lean ore dressing cost management and optimization method at home and abroad and proposing the difficulties and key issues in ore dressing cost system optimization.The cost optimization of ore dressing production process is not a simple problem for cost index management and control,is a multi-objective,non-linear,and dynamic optimization problem which closely coupled with production processes.Present methods are difficult to consider multifactors and their complex relationship.So,building a cost system and coordination optimization model for the production process is crucial to systemic cost optimization.2)Researching the dynamic cost system optimization processes in lean iron ore dressing and forming the cost system optimization mechanisms in lean iron ore dressing.According to the theory of metal quantity conservation and quality-cost management and aiming at cost management structure and process.this research analyzed the relationship between ore dressing cost index and production index;formed spatio-temporal multi-dimensional digital modeling method for ore dressing production system;established distributed spatio-temporal data model for progressive production process including cost index,production index,operation index,and control parameter;and built cost system optimization mechanism for lean iron ore dressing based on cost optimization and production process coordinative optimization.3)Researching the optimization method for selecting lean ore raw materials and building the optimization model of ore raw materials selection considering the ore processability.The selection of ore raw materials directly affects the management and cost of ore dressing production.In this research,the relationship between the cost and production index in mining and ore dressing was established,and the optimization method of ore raw materials selection considering the ore processability was formed.Combined with the mining and beneficiation process,the optimization method of ore raw materials selection and the improved Particle Swarm Optimization(ADIPSO)algorithms were studied,and the optimization model considering the ore processability was established.4)Building cost optimization model for production management and control based on cost optimization mechanisms in lean ore dressing processes.Based on "the most economic management" system optimization idea and spatio-temporal data model,spatio-temporal data mode for scenarios and its index adjustment method were established using spatio-temporal data and knowledge accumulating in ore dressing production;the prediction algorithm for cost index,production index,operation index,and control parameter was studied by deep learning;the management and coordinative optimization model of management and cost system for ore dressing production process was built based on spatiotemporal reasoning.The research on the application of related theories and methods was based on the QDS iron ore dressing production.The effectiveness and practical value of the proposed coordination optimization mechanism,algorithm,and model about ore production cost index,production index,operation index,and control parameter were verified using the actual production data.