Research On Predictive Control Algorithm Of Cement Rotary Kiln System Based On IELM-ARMAX Model

The cement industry is the basic industry of the construction of social facilities,the quality and output of cement is related to the construction of the firm and construction progress,therefore,the production of cement is very important to the current society.The cement rotary kiln is the core equipment of cement production line,and the cement rotary kiln system is the most important part of cement production,what's more,the core variable of cement rotary kiln system is the temperature of cement clinker burning zone.Based on researching of the cement rotary kiln burning zone temperature of related parameters,and combining the steady and dynamic model,establish IELM-ARMAX combination model of cement rotary kiln firing system for data modeling,infer the model predictive output,correct the predicted output error by the sequence of two quadratic programming method,and use online rolling optimized forecast input to achieve the optimal solution,to achieve the goal of cement burning system predictive control,through simulation experiments to analyze the effect of predict control,and the results show the feasibility of the algorithm.Specific work as follows:Firstly,in view of the characteristics of the cement rotary kiln firing system,identify the steady and dynamic characteristics of the cement firing system by using the IELM steady model and the RELS of ARMAX dynamic model method.The IELM method improves the disadvantages of traditional ELM method,by gradually increasing the hidden layer nodes,and improves the identification accuracy and ensure the rapid performance of the method;RELS can be employed to identify the complex cement burning system with colored noise,to avoid the disadvantages of the traditional least squares method.Secondly,this paper uses one parallel mode by connecting static and dynamic model,and then uses the steady-state gain parameters of steady state IELM model to adjust the dynamic gain of dynamic ARMAX model,completed the steady and dynamic model organically,construct IELM-ARMAX model predictive control algorithm which can comprehensively reflect the stability and dynamic characteristics of the complexmultivariable system,and achieve the output function of the model predictive control,then using the SQP method to complete the on-line optimization of the model predictive control algorithm.Finally,in view of the high temperature,high pressure,dust environment environment of cement burning system,it is hardly to direct measurement of temperature of cement calcining,this paper put forward an indirect method which uses oxygen and nitrogen oxide of the smoke chamber,direct relationship parameters with the temperature,to react the cement clinker burning temperature,and uses the IELM steady state model and dynamic model of ARMAX to identify system respective for the cement rotary kiln,then realize the predictive control of the cement rotary kiln firing system by using the IELM-ARMAX model predictive control algorithm.