Reaserch On Cement Grinding System Modeling And It’s Control Method

Cement is widely used in national defense construction, water conservancy project,construction industry, so the total amount of China’s cement production is huge. Optimizingof the production process is always under studying. As a result, how to improve the quantityand quality of cement, reduce the power consumption of cement production and reduce theenvironmental pollution caused by the cement production process is all cement productionenterprises’ common target. Every cement plant is improving in these aspects.In order to reduce energy consumption, most cement plants use dry rotary kiln method toproduce cement. The production process of this method can be simply summarized as "twogrinding and one burning". First, the raw materials are grinded in a ball mill, and then thegrinded materials are burned in the rotary kiln. Finally, the burned material combined withother raw materials go into the ball mill for further grinding, and after the grinding process thequalified material is cement. In the whole process, the grinding period consumes the mostelectricity, and grinding costs accounted for the total cost of production of about35%.Now, most cement plants use distributed control system to improve the degree ofautomation in the production process. Although the effect are greatly improved comparedwith the previous manual control, but this system still requires the experience of operator.Operators adjust the equipment in the central control room according to the change ofworking conditions of the production. Due to the uneven oprator’s operation level, the cementquality fluctuats greatly. What’s worse, the equipment’s spoilage are appeared and the powerconsumption are increased.This paper is based on the process and characteristics of grinding system of productionline in Pingyin Shandong Cement Plant. The current of out-mill lifting machine is chosen asinput while the speed of circulating fan is chosen as output. The model of this grinding systemis identified by the weighted least square method on-line. The qualified model is applied tothe predictive controller, and the speed of circulating fan is calculated. At the same time,another speed of circulating fan is calculated by fuzzy controller using the level of weightedbin and its’ change rate. Using appropriate weighted algorithm, a reasonable weighted value iscalculated so the final speed of circulating fan for control use is got by the combination of the two speed. By using predictive control and fuzzy control, the automatic control of thegrinding system is achieved. The load of the ball mill is stabilized, and the electricity of eachton of cement is reduced. In addition, the operator’s labor intensity is much lower and theanticipated effect is achieved.All the algorithms are based on Visual C++programming, and Matlab hybrid technologyis used, through the OPC interface the data exchange between the software and the distributedcontrol system is achieved. The software was successfully put into operation in Pingyinproduction line, and successfully extended to Rizhao Port Cement production line. Both ofthem achieve good control performance.Finally, on the basis of the completed work, some questions for further research andexploration are put forward.