关键词: 大型旋转机械, 故障诊断, 模式识别, 时间序列法, 专家系统

Abstract: The fundamental principle and advantage of a fault diagnosis method using the TSA(Time series analysis) are introduced. The information distance function for rotating machines is established with the AR models. This method is applied in the pattern recognition rule database for fault diagnosis expert system of rotating machine. Furthermore it is validated in the testing equipment.

Key words: Expert system, Fault diagnosis, Large scale rotating machine, Pattern recognition, Time series analysis