Research On Fault Mechanism Of Rolling Bearing And Intelligent Detection Technology

Manufacturing is the principal part of a country's economic development.Particularly high-end equipment manufacturing is closely related to social economy and national security.With the wide application of rolling bearing,the development of manufacturing is facing a series of problems,such as equipment maintenance,mechanical maintenance,life prediction,state identification and so on.The intelligent detection technology for different reasons,different representations and different degrees of mechanical failures has become the hot spot and difficulty in the current research.In this paper,the rolling bearing is taken as the object,based on the fault mechanism,the signal correlation technology is studied.The main content as follow:(1)The structural characteristics,classification and fault types of rolling bearing are described.From the view of the fault mechanism,the kinematics analysis of the rolling bearing is carried out.Then,the dynamic,fault load and fault signal of the rolling bearing are modeled.and the influence of fault location on signal is analyzed.(2)According to the knowledge of signal processing and machine learning,detection technologies based on AR-spectral-entropy and dictionary-reinforcement-learning are proposed.The former uses the AR model denoising,and the spectral entropy is used to select the best envelope in the frequency domain.The latter uses dictionary learning to preprocess,and introduces the AdaBoost-BP-NN to realize intelligent detection.The simulation and test results are significant.The technologies have broad prospect.(3)The intelligent detection platform is composed of the structure design,control and software development.The main contents includes the selection of motor and rolling bearing,the installation of sensors,signal acquisition,system programming,data visualization,numerical calculation and interface design.Fault mechanism of rolling bearing and proposed technologies are applied to the rolling bearing intelligent detection platform.The test results prove the effectiveness of the platform.