Single Sensor Information Fusion for Local Fault Prediction of Large Rotating Machinery

Data process of large rotating machinery is in line with basic features of information fusion. A framework of fault diagnosis and prediction based on sensor information fusion is built. An improved extracting method of features is used to deal with the information fusion of single sensor, which raises the calculation efficiency and precision. The local fault prediction process is presented, and the fault deterioration trend is judged on the basis of dynamic weighted method. Actual example of Beijing Yanshan Petrochemical Co. shows the correction of conclusion.