Study On Modeling And Predictive Control Of Combined Cement Crinding System

Cement is an essential and basic raw material for infrastructure and economic development. One of the important links is cement combined grinding in In its production. The link is a complex system which is made up of weighing bin, roller press, broken machine?v-separator?, ball mill, separator, exhaust fan and so on, and weighing bin and cement particle size have an important influence on production stability and quality of cement. Therefore, this paper focuses on weighing bin and cement particle size the two emphases, and we work on cement grinding model and predictive control research. Specific research works are as follows:?1? Aiming at the modeling problem of combined grinding system with cement particle size, this paper proposed a modeling method of combined grinding particle size which is divided into different conditions. According to the combined grinding process and on-line particle size analyzer, we analysis the interrelationship of key variables, and establish cement particle size conditions modes?two typical working conditions intervals? through the historical data; moving average filtering method is adopted to reduce the influence of historical data noise on Modeling; Aiming at typical working conditions 1, regression analysis method is used to establish particle size model with multiple-input and single-output; Aiming at typical working conditions 2, least square support vector machine is adopted to corresponding modeling; The simulation results show that the above established model based on the cement particle size conditions mode template can well describe the dynamic change process of cement particle size.?2? In order to realize the stable control of combined grinding's weighing bin, we put forward a modeling and internal model control method based on neural network extreme learning machine?ELMNN?. Based on analysis of relationship between combined grinding system process and variable, feed rate is determined as major factors which influence material level of weighing bin, moving average filter is adopted to reduce the noise of the data. ELMNN is adopted to establish internal model of weighing bin, internal model controller of weighing bin is designed by Taylor series, and the stability of the closed loop system is analyzed. The simulation results show that the modeling method and the controller we proposed can achieve stability control of material level of weighing bin.?3? Aiming at the stability control problem of cement combined grinding particle size with nonlinear characteristic?<45 sieve residue?, we proposed a generalized predictive granularity control method based on model. Based on the typical conditions 1's model established in?1?. Generalized predictive controller is designed through controlled auto-regressive integrated moving-average?CARIMA? model and long-time Optimization performance index. With the aid of the particle size closed loop transfer function, the closed-loop system is converted to particle size internal model structure form, and the stability of the closed loop system is analyzed. Simulation results demonstrate the effectiveness of the proposed method.?4? Based on the research results of?1? ³, the automatic control software platform of cement combined grinding system is proposed combining expert system, Bang-Bang control, internal model control and generalized predictive control algorithm. Engineering application shows that the platform has good operating results.