EMD分解分形理论和RBF神经网络相结合的轴承智能故障诊断研究.pdfVIP

EMD RBF 1 1 2 1 1. 210007 2. 518001 EMD EMD IMF IMF RBF EMD RBF EMD RBF TH17 A (2012) 11 0010 06 EMD decomposition, fractal theory and RBF neural network fault diagnosis of bearing intelligence research ZHANG Mei-jun1 1 2 1 CHEN Hao CAO Qin WANG Chuang ( 1.Institute of Field Engineering, PLA University of Science and Technology, Nanjing 210007, China; 2.Civil Contruction Office, Guoxin Securities Co. Ltd, Shenzhen 518001, China ) Abstract The diagnosis of accurate and fast solving mechanical equipment and production is very important. In order to improve the veracity of the intelligent diagnosis of rolling bearing, this paper will use the organic combination of EMD decomposition, fractal theory and the neural network, get the IMF component, and then draw the IMF component connection points bi-logarithm charts and draw the signal correlation dimension. Use the correlation dimension and RBF neural network to bearing state of the classification and recognition, to achieve the purpose of intelligent diagnosis. Example analysis shows that EMD decomposition, fractal theary and RBF neural network can effectively reduce the nonlinear component of fault signal interference, and accurately identify the rolling bearing fault type. The combination of the three proved effective and feasible intelligent fault diagnosis. Key words fault diagnosis rolling bearing EMD decomposi