The Application Of Predictive Control In Cement Raw Meal’s Quality Control

Cement is an important building material,the cement raw meal is the base of producing cement products,and has a great influence on the quality of cement products.Most cement enterprises in China still rely on manual means to produce cement raw meal,these methods have low accuracy and poor real-time performance,It is difficult to obtain high quality cement raw meal,achieving the automation of cement raw meal’s production and improving the quality of cement products is an important issue to be solved.The near-infrared element analyzer on-line can detect the content of various kinds of oxide in cement raw meal in real time,which is help to realize the automation of cement raw meal’s production.However,the production process of cement raw meal has the characteristics of delay,inertia and disturbance,so it is difficult to control,and it is of great significance to study the quality control of cement raw meal.This topic’s research background is one 5000t/d cement product line in Shandong province of China,we according to the production process of cement raw meal,combining with near infrared element analyzer on-line and production site situation,propose the use of predictive control and static model respectively adjustment different work condition,improving the quality control software for cement raw meal,the main contents are as follows:(1)Development of predictive control module for cement raw meal’s quality control system.When the KH of cement raw meal needs to be adjusted and the other two rate values need not to be adjusted,changing raw materials’ ratio of the limestone and the mudstone can achieve adjusting goal.In this case,two kinds of raw materials’ ratio will be changed,and the oxide’s content in cement raw meal mainly changes two types.Because of the cement plant where we make test attach great importance to the KH,and the chemical composition of limestone is unstable,needing to adjust the ratio of limestone frequently,and the application effect of static model is not good,but it is convenient for prediction model’s identify and predictive control’s application.In this paper,we use the ratio’s change of raw materials as inputs,the variation of CaO and SiO2’s content in cement raw meal as outputs to identify model,and verifying the model;then predictive controller is designed based on the model,and parameters are decisived by simulation.(2)Development of correct control module for cement raw meal’s quality control system When the rate of iron or silicon of cement raw meal are not qualified,the raw material’s ratio of sandstone or iron tailings should be adjusted in order to achieve the purpose of regulation.In this case,the number of input variables and output variables is large and the coupling between variables are strong,however,the chemical composition of sandstone and iron tailings is stable,and the relationship between the oxide content’s change in raw meal and the ratio’s change can be calculated by the static former.In this paper,we combine static model with expert rule to control the working condition,It is a modified control for the quality control system of cement raw meal,In addition,we put forward the method of modifying the composition of raw material based on the principle of least square method.At last,we simulate the control method and get corresponding control parameters.(3)Realization of quality control system of cement raw meal.In the production field,we based on the element analyzer on-line to collect real-time data,realizing the data exchange between hardware system and database by OPC,using cement raw material quality control software to identify the various conditions and regulate corresponding working situation.This system can realize the automation control of cement raw meal’s production and improve the quality of cement raw meal.In this paper,we count and analyse the production data from site,the results show that the quality control system of cement raw meal can automatically and stably work for a long time,and the control effect is better than that of manual control.