Predictive Control Of Cement Combined Grinding Based On Typical Working Conditions

As the key industry of basic raw materials,cement industry plays an important role in the national economic construction.Cement combined grinding as the most important part of cement production determines the quality and yield of cement.And the high consumption and low efficiency of cement combined grinding system determines it's necessary to save energy.The main part is made up of weighing warehouse,rolling machine,broken machine(v-separator),ball mill,circulating fan and so on.There is a strong coupling,large time delay,nonlinear problem.The production process presents the characteristics of the fluctuation range of process parameters and working conditions changing,complex operation means,so it's of great significance of stable and efficient control of the production process.This paper makes a deep research on the working condition template division,modeling and controller design of the combined grinding system.The main contents of this thesis are as follows.(1)In view of the cement combined grinding system,the working condition template is established.According to the mechanism of combined grinding process and experience of outstanding field worker,combined with the production line of a cement plant,the actual operation data,analyzed the main factors that weighing warehouse material level and ball mill electric current,3D curve,defined the interaction relationship between the cement combined grinding variables,reasonable template for the combined grinding is established.In order to establish the combined grinding model based on typical working conditions and to establish the foundation for the precise control of the grinding model.(2)According to the condition template established in(1),taking the pre-grinding section of the roller press as an example,the feed quantity,the former separator speed and the circulating fan speed are selected as the input variables needed for modeling,and weighing warehouse material level as an output variable.After the field parameters data are collected,the model data are obtained by rolling filter.For the cement combined pre-grinding system,the model is established according to the different working conditions.For the typical case 1 and the typical case 4,the least squares support vector machine(LS_SVM)is used for the corresponding modeling.For the typical case 2 and the condition 3,the regression analysis method is used to build the model.The degree function combines the sub-models to the final model.The simulation results show that the proposed model has high precision and provide the basis for the stable control of the steady weighing warehouse.(3)Aiming at the problem of stability control of weighing warehouse material level with non-linearity,a generalized predictive weighing warehouse material level control method based on model is given.The generalized predictor designed in this paper is based on the controlled autoregressive integral moving average model(CARIMA).Based on the model established in(2),the condition 2 and the condition 3 are linearized,and the generalized predictor can be established directly by the CARIMA model;the working condition 1 and 4 is nonlinear,but the system of the working range is vary small and the equilibrium point of the nonlinearity is weak enough,the linear model of the controlled system can be obtained off-line at each sampling time and the linear fitting model can be used to obtain a more accurate fitting result,and then establish a generalized predictor according to the CARIMA model.MATLAB simulation results show that the generalized predictive control is more stable and better than PID control.