Study On Optimization Of Cement RAW Meal Proportioning Based On Clinker Quality Prediction

Cement manufacturing industry is a typical process industry,and any link in its production process will have an impact on the follow-up link.Because of the continuity of cement production itself,a higher degree of comprehensive automation is the necessary prerequisite for stable,efficient and low consumption cement production.At present,the automation level of raw cement quality control in China is low,and most cement plants still stay at the level of manual control or semi-automatic control.The three-rate target values of raw meal proportioning(limestone saturation ratio coefficient KH,silicic acid ratio SM,aluminium oxide ratio IM)are given by laboratory staff according to the clinker sample test results extracted from the site and combined with manual experience;the raw meal proportioning is given by the operator of the central control room according to the abrasive raw meal sample test results,and the rate target values given by the laboratory.Adjustment based on manual experience.It can be seen that at present,the whole raw meal proportioning control process is mostly completed manually,and human factors have great interference.In addition,due to the irregular mining in China,the fluctuation of raw material composition and the instability of cutting mechanism are the characteristics.It is easy to cause such problems as untimely adjustment of target value of rate value,low accuracy of raw material ratio and long lag time.Therefore,the ratio of raw materials is often adjusted several times before it works.The adjustment lasts as long as 3 to 5 hours.At this time,hundreds of tons of unqualified raw materials may have been produced,which will inevitably affect the calcination of rotary kiln and reduce the output and quality of clinker.In view of the above problems,this paper combines the application of near infrared online analyzer in automatic control of raw meal proportioning,aiming at optimizing the target value of three-rate value of raw meal proportioning,coordinating raw meal proportioning and clinker calcination,stabilizing the quality of raw cement meal,thereby reducing the fluctuation of kiln working conditions and improving the quality of cement clinker.The detailed work of this paper is as follows:(1)Research on clinker quality prediction model.In view of the problem that laboratory personnel adjust the target value of raw meal proportioning rate according to off-line clinker quality,and the feedback time is too long to adjust in time to meet the demand of on-line analysis of raw meal proportioning,this paper uses RBF neural network algorithm for clinker quality prediction modeling.Based on the clinker quality calcination process,the input and output variables of the model are selected to determine the time matching scheme of variable data extraction and data preprocessing;the activation function of radial basis function neural network is selected to gradually obtain the model parameters,and the clinker KH prediction model based on RBF neural network is established.(2)Working condition division of rotary kiln.The clinker calcination process has the characteristics of non-linearity,multi-variable and strong coupling.The working conditions of the rotary kiln are also different in the material calcination process in the rotary kiln.This paper mainly identifies the fluctuation of kiln working conditions caused by the change of raw meal quality.Firstly,the known raw meal quality historical data are extracted.According to the lag time pair rule,the firing parameters data of rotary kiln under this material are extracted.Then,K-means clustering method is used to cluster,and the clustering results are analyzed.Four kinds of kiln working conditions are divided according to the operation experience of field knowledge workers.Corresponding adjustment schemes are given for different working conditions.(3)Development of raw meal batching optimization software based on framework/rule expert system.According to the above research content,the target value of raw meal proportioning rate is given by synthesizing the firing condition in kiln,off-line data clinker rate value and predicted clinker rate value,and an expert rule system suitable for field application is built.Finally,according to the production demand and the use of on-line analyzer,the optimization software of raw meal proportioning is developed.Finally,combined with the application of near infrared on-line analyzer in raw meal batching,the raw meal batching optimization software was put into operation on site.The results show that the research results of this subject have great significance in guiding the quality control of cement raw meal,stabilizing raw meal quality and improving clinker quality and production efficiency.