A New Method of Dynamic Prediction on Remaining Useful Life of Barrel

摘要

To fully use the performance degradation similarity of the congeneric barrels, for the traits of nonlinear and small sample during the performance degradation process of a barrel, a dynamic prediction model on remaining useful life (RUL) of barrel is proposed, on the basis of fuzzy c-means theory and support vector regression optimized by the optimum model of particle swarm optimization. Lastly, the effectiveness and superiority of this proposed model have been validated by the case of a barrel compared with the improved Verhulst gray theory. In conclusion, a new method is obtained for the prediction on remaining useful life of barrel.