Kalman filter is applied to equipment fault prediction. In practical application, considering the influence of noise and disturbance on the measurement results, the Kalman filter can be used to predict a group of data containing measured noise while analyzing the measured noise; the Kalman filter is easy to design computer program, can update and process the field data in real time, and is easy to implement and combine with other algorithms. Through the discussion of the key methods of Kalman filtering and the analysis of examples, the results show that Kalman filtering has a good application prospect in equipment fault prediction.