Design And Implementation Of Intelligent Coal Blending System

In the actual coking production process,under the premise of smooth coking conditions and stable quality of single coal,the coke quality is mainly affected by the quality of coal blending,and the quality of coal blending is determined by the quality of single coal and coal blending ratio,so the main factor affecting coke quality is coal blending ratio.Based on the actual coking production data of the coking plant,this paper explores the coal-coke-working condition data relationship,and proposes a coal blending ratio calculation method,which can accurately calculate the coal blending ratio,improve the coal blending efficiency,and save the production cost.The research work of the thesis mainly includes:1.Starting from the production process of coking coal blending,the relationship between single coal and blended coal is analyzed,and the quality calculation model of blended coal is constructed by linear weighting.The factors that affect coke quality are analyzed in detail,including blended coal quality and coking conditions.BP neural network is used to build the actual coke quality prediction model,and the prediction accuracy of the prediction model with and without conditions is compared and analyzed.In order to improve the prediction accuracy of neural network,particle swarm optimization(PSO)is adopted to further optimize the coke quality prediction model constructed by BP neural network,and the prediction accuracy of BP neural network and particle swarm optimization(PSO)BP neural network is compared and analyzed.The results show that PSO BP neural network improves the prediction accuracy of coke quality.2.Aiming at the shortcomings of traditional manual coal blending,such as low work efficiency and large quality fluctuation,large dataanalysis is carried out on the actual coking production process.Coking production process data affecting coke quality are analyzed and extracted from coking enterprise quality reports and production process reports.Accurate matching of coal quality-coking conditions-coke quality data is completed.Based on required coke quality parameters,given working condition data and single coal quality parameters,an intelligent coal blending method is proposed considering production cost,inventory and other related constraints,and a coal blending ratio calculation model is established,through which the coal blending plan can be quickly calculated.The regression analysis is carried out on actual coke-coal-working condition data,expert experience is sorted out,and expert knowledge base for optimizing coal blending ratio is established.3.Under the.NET development platform,the application software of intelligent coal blending system is developed with C# and SQL database as development tools.The system runs stably in coking plant.Enterprise feedback results show that the intelligent coal blending system can meet the actual production needs of coking plants.The designed intelligent coal blending algorithm can quickly and accurately calculate the coal blending ratio,save costs,reduce production risks and improve the prediction accuracy of coke quality.