A real-time residual life prediction method based on adaptive bandwidth kernel density estimation is proposed for multi-component systems with complex stochastic dependence. Firstly, the dependence modeling of the components with stochastic dependence effect of continuous degradation is carried out. Secondly, a degradation model of conditional kernel density estimation based on adaptive bandwidth is constructed to obtain the corresponding probability density function. Finally, a prediction model of the remaining life of the component is established to obtain the residual life of components. The accuracy and effectiveness of the proposed method are verified by experiments on a helical gearbox test-bed.
|Translated title of the contribution||Remaining life prediction of multi-component system influenced by stochastic dependence|
|Original language||Chinese (Traditional)|
|Pages (from-to)||299-307 and 324|
|Number of pages||10|
|Journal||Zhendong yu Chongji/Journal of Vibration and Shock|
|Publication status||Published - 15 Nov 2022|