Prediction Method Of Equipment Reliability Based On Kalman Filter After A Deteriorating Failure Occurrence

The deteriorating failure is universal in practice, which is caused in the situation that the equipment components gradually deteriorate to a certain degree. The research on the reliability prediction for equipments after deteriorating failure occurrened (RPEDF) is very important for avoiding catastrophic accidents, arranging rational production plan and reducing the maintenance cost. RPEDF belongs to predictive maintenance, which has been a hot issue. At present, reservice life prediction and accelerated degradation testing are often adopted, the former is a indirect method, and the cost of the latter is high and there may be a great deviation with the actual. Therefore, based on the historical data of the equipment and selfish condition, we adopt kalman filter to forecast the reliability. The main content is as follows:Firstly, the thesis makes a thorough review of research on the theory of forecast methods for equipment fault and RPEDF at both home and abroad. It also summarizes the theory of reliability and kalman filter. Secondly, in accordance with the change of deteriorating values, the reliability prediction model is established to forecast the average value of degradation in the future and further the reliability of the future is predicted. The method is designed by kalman filter and the detailed algorithmic flow of reliability forecasting is also offered. In the method, the average value of degradation in the future is predicted by kalman filter, the failure threshold is defined by Delphi methods, and also a suboptimal fading kalman filter is introduced for improving the ability to track the sudden change and the adaptability of the method. Finally, the simulation examples are used to validate the effectiveness of the proposed method.