A Particle-Filtering Approach for Remaining Useful Life Estimation of Wind Turbine Gearbox

摘要: Machine failure prognostic is concerned with the generation of long term predictions and estimation probability density function remaining useful life. For this we propose a framework for data-driven prediction RUL. To solve problem lacking direct condition information in predicting equipment residual life (RUL), particle-filtering model built equipment's RUL indirect information, which easy to get .This paper introduces modeling approach lifetime wind turbine gearbox based on SCADA system monitoring. Data from were used validate proposed methodology. The outcome shows that PF method has better effect prediction.Finally, verified through on-site data collection. It practical value A new way state recognition complex provided.