Real-time reliability prediction of mechanical equipment based on particle filter

This paper presents a particle filtering method for real-time reliability prediction. In order to describe running status of equipment, a model of status was created by making use of prior knowledge. Then, through changing newly-obtained measurement data by weighted probability, model of observation is set up, so as to correct previous prediction and ensure nicety of following prediction. The proposed method is not restricted to any contribution. It can effectively deal with non-linear, non-placidity and non-Gaussian problem of prediction and evaluation in equipment status. Experiment result shows that the proposed algorithm is effective and robust.