Modeling And Toolbox Development Based On SVM For Rotary Drying

In recent years, the high energy consumption and heavy pollutionphenomenon in the rotary kiln drying process of Metallurgical industry isparticularly prominent, which has severely restricted the economic de-velopment of the industry, so the enterprise and the government begin toattach great importance to the study of energy conservation and emissionsreduction. Drying kiln which is regarded as the main body of the rotarydrying process has many characteristics, such as large delay, nonlinear,time-varying, multivariate parameters, and it makes the establishment ofrotary kiln drying process model have higher requirements. Therefore, theresearch on optimal model of rotary kiln drying process is of great signi-ficance to improve the efficiency of enterprises.In this paper, taking a smelter rotary kiln as the background, throughin-depth analysis of technological process and control requirements ofrotary drying process, we have mastered the key factors affecting thedrying process modeling. Then the establishment of the mechanism mod-el and the hybrid model of rotary drying process, the improvement of thehybrid model accuracy are studied. And one modeling toolbox based onMatlab/GUI is developed which is on the basis of the summary and ref-erence basis of many previous modeling methods and intelligent optimi-zation methods. The main research and innovative achievements of thispaper are as follows:Firstly, the complex process and control requirements of kiln dryingprocess are discussed in this paper, and some drying rate experiment dataare obtained by adding water humidifying method. The drying rate is ob-tained by using traditional methods, and the mechanism model is builtbased on the mass and energy conservation of rotary kiln and the simula-tion analysis is carried on at last.Secondly, aiming at the problem of obtaining the drying rate whichis the key parameter in the rotary kiln drying process modeling, a fuzzymodel of drying rate based on SVR is established in this paper, whichlays the foundation for the establishment of the rotary drying process hy- brid model.Then, based on the traditional mechanism model of rotary dryingprocess, the hybrid model based on support vector machine is designed,and the residual compensation method for the hybrid model is proposedby utilizing improved online SVR, then the simulation results show thegood effect of this method.Finally, this paper analyses the situation and demand of drying mod-eling toolbox, by selecting appropriate toolbox development tools, andaccording to demand analysis the overall structure of toolbox is designed.Then the development of drying modeling toolbox based on support vec-tor machine is completed by using the MATLAB GUI function, then itsvalidity and practicability is verified by application examples.