Research And Development Of Predictive Control System For Cement Combined Grinding Mill Load

Because of the rich raw materials and good performance,cement is widely used in construction projects and various infrastructure projects,it has already become an indispensable construction raw material.The ball mill is the most important equipment in the cement grinding process.the total amount of materials in the mill are called mill load,it will greatly affect the running state of ball mill.High load,resulting in a weak grinding effect of the mill,so that the quality of finished products;low load,not only cause energy waste,and easy to damage the equipment.Therefore,controlling the mill load within a reasonable range is conducive to stabilizing product quality and improving enterprise benefits.This paper gives a detailed introduction to the cement combined grinding process and major equipment,many factors affecting the change of mill load were analyzed,the difficulties of mill load modeling and control were clarified,and the key variables of mill load were determined based on the actual production situation and production data.Aiming at the problem of ball mill load modeling,a subspace identification method is introduced,some geometric and linear algebra methods are used to obtain the signal subspace of the system.furthermore,the system matrix is obtained by using the least square method,thus,a state space model is established with the speed of circulating fan as input and the current of cement grinding mill elevator as output.In order to realize automatic control of mill load,a direct subspace predictive control method with simple structure and easy implementation is considered on the basis of subspace identification.In the traditional model predictive control framework,the subspace linear predictive equation is used as the predictive model,and then the controller is designed;in view of the fact that there are constraints in actual production,a subspace predictive controller with constraints is designed,the relevant constraints were processed,the system input is obtained by solving a quadratic programming problem,finally,the effective control of the mill load is achieved.Considering application in engineering,MATLAB and C# hybrid programming method was adopted to develop the predictive control algorithm based on subspace to ball mill load forecasting control system,and aiming at the situation in which mill load fluctuations caused by the instability of raw material grindability in actual production,using the feedforward compensation method based on expert rules,and finally,the simulation test of the control software verifies the effectiveness of the system.