Research On The Condition Of Blast Furnace Based On PCA And SPC

The blast furnace smelting process has always occupied an leading position in the world’s iron smelting process.Because the internal environment of the blast furnace smelting process is complicated and changeable,and the existing technology does not fully understand the evolution mechanism of the furnace conditions.Therefore,a scientific and objective method of judging furnace conditions is an urgent issue for iron smelters.In the last few years,in the wake of developments of information technology,blast furnaces have accumulated a large amount of actual production data during long-term operation,which provides a good foundation for theoretical research.In view of the complex and changeable blast furnace smelting process,there are many factors affecting the condition of the blast furnace and the relationship is complicated.The work of this paper is as follows:(1)In order to ensure the stability of the blast furnace condition,a detection model of abnormal condition of blast furnace based on principal component analysis(PCA)and statistical process control(SPC)was proposed.This paper first uses the principal component analysis algorithm to cluster the historical data of the actual production of the blast furnace,reduces the high-dimensional data of the blast furnace itself to low-dimensional.Secondly,using the multivariate control chart based on the T2 statistic to analyze the new variables Monitoring.And then establish a blast furnace abnormal furnace condition monitoring model.The simulation verification of offline historical data shows that the algorithm realizes the detection of abnormal furnace conditions of the blast furnace and verifies the validity of the model.It demonstrates the possibility of applying SPC algorithm to online real-time monitoring of blast furnace smelting process,and provides a new idea for iron and steel enterprises to grasp the quality of molten iron.(2)Building a Spark-based big data analysis platform.Applying the above model to the Spark Stream streaming data framework through the Py Spark tool to realize the algorithm module of data cleaning,data mining,process monitoring,fault diagnosis and other functions.Using data instead of experience to help blast furnace operators observe the changes in blast furnace conditions in real time and ensure that the blast furnace is in a stable operating state.