The Fuzzy Control For The Cement Combined Gringing System

Cement is an important material in the construction industry, the production process is one of the more important aspects of the joint grinding, and its stable operation can ensure the quality of cement, production and so on. Therefore, the research on intelligent control of combined grinding system is very meaningful. Combined grinding system is a complex system with nonlinear, large delay and strong coupling characteristics. It is very difficult to accurately describe the production process of grinding system. This paper puts forward the fuzzy control of combined grinding system. The specific research contents are as follows:(1) For accurately describing the change process of the combined grinding system, establishing the T-S model, the key variables are determined by analyzing the relationship of the data. Analyzing the technology of combined grinding system, combined with operator experience, initially to determine the key variables of the mill load and the warehouse level; analysis of the trend of the historical data line graph, determine the key variables of T-S models. MATLAB simulation results show that, the key variables that affect the mill load is the rotational speed of the classifier, and affecting the warehouse level is the total given. In this paper, the speed of the classifier can control mill load, total given can control the warehouse level, the T-S models will be established between this two relations.(2) For the realization of T-S models of mill load and steady flow Warehouse, the former based on fuzzy c-means clustering, the second component based on least square algorithm. Collecting the historical data of the main variables; the mean filtering method is using for noise reduction in the selected data, processed data then can analysis by fuzzy c-means clustering algorithm, cluster centers and the membership degree matrix can be got. According to the results through clustering algorithm, we can identify the parameters of T-S model by using least squares multiplication algorithm. Using MATLAB software, according to the data fitting and data verification of simulation results shows that the T-S model can reaction very good the mill load and stable chamber material change process.(3) A generalized predictive control algorithm based on T-S model is proposed to realize the fuzzy control of the combined grinding system. T-S model is composed of a plurality of sub models. Therefore, in this paper, the generalized predictive controller of the sub models are calculated simultaneously, according to the current operating point belongs to each cluster centers of the two norm obtained membership degree of each sub model; sub models of the calculated control output by fuzzy weighted to get the final output. MATLAB simulation results show that the generalized predictive control is better than the PID control, the control accuracy is higher, and the stability is better.(4) In order to realize the control software of combined grinding system, a method based on C# and MATLAB mixed programming will be put forward. The window procedure will be writing through the language of C#, and the control algorithm will be writing in the MATLAB and then a generate dynamic link library will be producing through the MATLAB; then the mixed programming will come true; the control software can connect DCS with OPC to control the combined grinding system with fuzzy control.