Research And Development Of Cement Vertical Mill Automatic Control System

The final step of the three grinding and one burning process in cement production is cement grinding.The stability of its core equipment,the self-supporting roller mill,is the main factor to ensure the quality of cement.However,considering that there are disadvantages such as difficulty in dynamic description and incomplete quality parameter detection in the grinding process,this will inevitably bring challenges to the automatic control of self-supporting roller mills.At the same time,most cement plants still use operators to manually adjust,resulting in low cement grinding efficiency and large fluctuations in cement quality.Therefore,this article aims at stabilizing the grinding process,improving cement quality,and reducing production energy consumption,combined with least squares algorithm,weighted least squares algorithm and particle swarm optimization algorithm to establish the feeding volume-intermediate bin weight,plate feeder the dynamic relationship between the speed-the pressure difference in the mill and the speed of the separator-the cement content of less than 45μm.The design process of the expert PID controller for cement grinding based on the self-supporting roller mill is proposed,and the cement based on the radial basis function neural network series is given.The calculation method of the command setting value such as particle size and mill current has been developed and applied to the actual industrial field.The main research contents are as follows:(Ⅰ)In order to ensure the stability of cement quality and improve the grinding efficiency,online particle size analysis equipment is introduced to monitor the particle size distribution in real time,the cement grinding process based on self-supporting roller mill is described,and various typical working conditions and key process parameters are analyzed,and proposed a self-supporting roller mill automatic control program.First,on the basis of the self-supporting roller mill cement process,an in-depth mechanism analysis is carried out;then,based on engineering realization,the characteristics of normal and abnormal typical working conditions are explained,and the foundation for the establishment of segmented dynamic models for different working conditions is laid;and finally;with the goal of achieving stable control of cement grinding,a self-supporting roller mill control plan is proposed,and the main state variables and control variables are analyzed.(Ⅱ)Based on the analysis of cement grinding conditions,the dynamic model of cement grinding based on self-supporting roller mill is established,and the PID control method of cement grinding experts is given.First,filter the field data with the help of the low-order difference method to reduce the influence of the coarse value on the dynamic relationship fitting;then,through the analysis of the feed amount-the internal pressure difference of the mill,the speed of the plate feeder the internal mill pressure difference and the separator-less than 45μm speed,the cement content of each control link has the characteristics of lag and inertia,and the least square method,weighted least square algorithm and particle swarm algorithm are used to identify and compare the models of each link;in this way,combined with PID control.Based on the advantages of expert control,an expert PID controller based on model identification is designed;finally,the effectiveness of the proposed control method is verified by simulation.(Ⅲ)In order to realize the independent control of cement grinding,based on the design of the controller,with the aid of the radial basis neural network,a calculation method of cement particle size and mill current setting value with the input of the weight of the intermediate silo and the internal pressure difference of the mill is proposed.First,the general idea of the calculation method of the setting value based on the radial basis neural network series is given,and the design process is briefly explained;then,the setting value adjustment network,the performance index prediction network,and the cement particle size working condition judgment are explained in detail.The specific design process of the module is finally verified by simulation to verify the effectiveness of the proposed method.(Ⅳ)On the basis of cement grinding dynamic model,expert PID controller,cement particle size and mill current setting value calculation,using Visual Basic,OPC communication technology and SQL Server technology,developed cement self-supporting roller mill monitoring with automatic control system software,and applied in engineering.