Optimization Of Burden Constituent And The End-point Prediction Model For50t Converter

Converter steelmaking is a sophisticated process due to high temperature, multi-phases and multi-reaction in the reaction system. When the converter is operated ininformal mode, the hittting rate of end-point temperature and carbon content in themolten steel is declined, then the melting cost should be increased and the quality of themolten steel should be declined. Therefore, on the basis of the technical and operatingcondition utilized during the production process, the controlling model of burdenconstituents optimization and the end-point prediction model established for medium andsmall converter operation possess.According to the production condition of50t converter used in a plant, theoptimization model of burden constituents has been established on the basis ofmechanism model and corrected by incremental model. BP neural network has been usedto set up the static prediction model. Process controlling model established on the basis ofthe operational experiences, and deoxidization, metallization and cost accounting modelshave been also built in this study.Based on these models established, burden constituents optimization and end-pointprediction system for50t converter has been constructed with VB software and SQLServer database. In this system, the predition model was incorporated with processcontrolling model to provide in-time pridiction for oxygen-supply and charging model,which could provide process pridiction for steelmaking and increase the hitting rate ofend-point prediction.The system has been used offline at the50t converter. It has been shown that theprediction hitting rate of end-point temperature and carben content in the molten steel is75%and78%, as the deviation between prediction and actual value is lower than15℃and0.02%, respectively. As the system is in operation offline, the burden constituentscould be optimized. Oxygen-supplying and charging model could provide importantindictions for process control. The addition quanntity of deoxidant and alloy could ensurethe end-point to meet the target.