Research On The Free Lime Content Prediction Model And Optimization Control Algorithm Of Cement Clinker Sintering Process

The problems of energy consumption,environmental pollution and quality improvement in cement clinker sintering process have become the focus of the cement industry.The clinker sintering process affects the content of free lime in cement clinker(fCaO)and the level of energy consumption directly,which is the key link of cement production line to save energy,reduce emission and improve the qualified rate.Considering the complex physical and chemical reaction features of cement clinker sintering process,it is important to study the online neural network structure,intelligent optimization algorithm and model control algorithm.This paper established the cement clinker fCaO content prediction model,and the research on cement clinker sintering process variable optimization and control algorithms were completed.These studies have not only important theoretical significance,but also a wide range of application prospects and huge social benefits.And these studies also have great practical value to promote the development of cement industry technology,improve energy efficiency,reduce air pollutant emissions and improve clinker quality.The specific researches work are as follows.Firstly,this paper introduces the technological process of cement clinker firing process,analyzes the physical and chemical reaction of cement clinker firing process,and gives the production process of cement clinker fCaO.Based on the mathematical description of gas-solid model,pulverized coal combustion model and radiation heat transfer model,the high-temperature gas-solid turbulent heat transfer model of cement clinker sintering process rotary kiln is established.The heat transfer characteristics of the rotary kiln were simulated by Fluent software.The influence of the burner parameters on the flow field distribution in the cold state and the influence of the relevant parameters on the temperature distribution during the pulverized coal combustion process are analyzed.Secondly,the kernel function method is introduced into the Extreme Learning Machine to construct the Multiple Kernel Extreme Learning Machine(MKELM),which combines three kinds of kernel functions.On the basis of MKELM online calculation process,the Cholesky Factorization Online Sequential Multiple Kernel Extreme Learning Machine(COS-MKELM)is proposed and tested by three regression datasets.Based on the technology and heat transfer characteristics of cement clinker firing process,the input variables of cement clinker fCaO content model are determined,and the cement clinker fCaO content prediction model is established.Experiments are carried out by data simulation verification experiment and simulation platform test experiment.Thirdly,two improved algorithms for optimizing the sintering process variable are proposed: Chaos Particle Swarm Optimization Based on Random Perturbation Mechanism algorithm and Ameliorated Quantum-behaved Particle Swarm Optimization Based on SQP Local Search algorithm.The benchmark functions are used to test the performance of the improved optimization algorithms.On the basic of cement clinker fCaO content prediction model,the input variables of the fCaO content prediction model are taken as the optimization variables of cement clinker sintering process,and the cement clinker sintering process variable optimization model is established.The sintering process variable optimization model is optimized by the proposed optimization algorithm.The optimal values of the cement clinker sintering process optimization variables are given through simulation experiments.Finally,the OMKELM-ARMAX combined modeling identification algorithm is deduced to establish the control combined model of cement clinker sintering process.The optimal values of the sintering process variable optimization variables are taken as the set values of the controlled variables.The predictive control algorithm based on combined model is proposed.And the optimal input increment expression of the cement clinker sintering process operational variables is given.The performance of the control combined model and predictive control algorithm of cement clinker sintering process are verified by data simulation verification experiment and simulation platform test experiment.