Operational reliability evaluation and prediction of rolling bearing based on isometric mapping and NoCuSa-LSSVM

摘要

In order to solve the problem of evaluation and prediction of rolling bearing operation reliability, a prediction method of rolling bearing operational reliability is proposed based on isometric mapping and nonhomogeneous cuckoo search-least squares support vector machine(NoCuSa-LSSVM). Firstly, isometric mapping (ISOMAP)algorithm is used to reduce the dimensionally of the high-dimensional collection composed of the characteristics of time domain, frequency domain and time-frequency domain of bearing vibration signal. Secondly, as the characteristics of the performance degradation state of the bearing, the integrated characteristic index establishes the reliability model of the rolling bearing through the logistic regression model. Finally, the NoCuSaLSVM model is used to predict the characteristics of bearing performance degradation state, and the results are embedded into the established reliability model, so as to obtain the prediction results of bearing operational reliability. The method proposed in this paper was verified by the whole life test data of rolling bearing from the university of Cincinnati.