The Study On Temperature Modeling And Control Of Cement Rotary Kiln

Cement is an indispensable material in the construction industry.With the accelerated pace of urbanization in our country,the demand for cement is increasing.High temperature is needed in raw material decomposition and sintering,therefore the cement production is a thermal process.And the burning zone temperature in cement rotary kiln determines the quality of clinker.For this reason,it is of great realistic importance and academic value to accurately obtain and control the temperature.Taking into account characteristics,nonlinear,large delay and multi variable,of the rotary kiln system,the traditional temperature measurement method can not effectively obtain the burning zone temperature.In this paper,method of data fitting is used to establish prediction model,and a study is done about the temperature control based on fuzzy control.First of all,the development status of new dry process cement industry was introduced,and summarized the domestic and foreign academic achievements on temperature prediction and control.By analyzing the production process of cement,nine variables related to the burning zone temperature were selected.According the gray correlation analysis method,the correlation degree of each variable and the burning zone temperature was calculated.Finally,this paper selected seven input variables.According to the characteristics of the rotary kiln system,this paper selected the least square support vector machine to establish the temperature model.And the parameters of this method were selected by the multi population genetic algorithm with strong robustness and not easy to fall into the local optimum.By comparing the simulation results and experimental data,the simulation results show the validity of the modeling method.Due to the influence of material components and the interference of the external environment,the condition of the cement rotary kiln system is time time-varying.In order to reflect the actual working conditions better and improve the accuracy of prediction,the online least squares support vector machine temperature prediction algorithm was proposed.Taking into account the problem that burning zone temperature is difficult to measure online,backend temperature was selected as the output variable.The algorithm updated the modeling data in real time and the new parameters were calculated by recursion,which realized online temperature prediction.Fuel feed rate was control variable of the temperature control system,and back-end temperature was controlled variable,which was easy to measure.Fuzzy control was chosen as the control method,and established fuzzy rules and query tables.Simulation results verify the effectiveness of the method.