Application of RS-SVM method in diagnosis of small sample fault of cement rotary kiln

2018 International Conference on Advanced Electronic Materials, Computers and Materials Engineering

材料科学;无线电电子学;计算机科学

Yang, Yiting^1 ; Mei, Canhua^1 ; Ouyang, Mingsan^2

Department of Information and Electrical Engineering, Huainan Vocational Technical College, Huainan

232001, China^1

College of Electrical and Information Engineering, Anhui University of Science and Technology, Huainan

232001, China^2

关键词: Cement rotary kiln;    Diagnosis methods;    Interference immunity;    Reduced data;    Small samples;    SVM theory;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/439/3/032002/pdf
DOI  :  10.1088/1757-899X/439/3/032002

来源: IOP

【 摘 要 】

The rough set (Hereinafter referred to as RS) and support vector machine (Hereinafter referred to as SVM) are combined and applied to the diagnosis of small sample fault of cement rotary kiln. In this paper, the basic knowledge of RS and SVM theory is introduced firstly. Then the knowledge of application of RS theory in cement rotary kiln fault is reduced. Lastly, SVM theory is utilized to train and classify the reduced data. This combined diagnosis method not only brings the advantages of two theories into full play, but also overcomes the limitation of SVM in identification of redundant information and useful information, effectively reduces the space dimensionality of input information of SVM and makes up the disadvantages of sensitive to noise of input information and poor interference immunity of RS theory. Therefore, the efficiency and accuracy of diagnosis are improved effectively.