Time Synchronous Averaging Based on Cross-power Spectrum

Abstract Periodic components are of great significance for fault diagnosis and health monitoring rotating machinery. Time synchronous averaging is an effective convenient technique extracting those components. However, the performance time seriously limited when separate segments poorly synchronized. This paper proposes a new method capable periodic without external reference accurate period to solve this problem. With approach, phase detection compensation eliminate all segments' differences, which enables be well The effectiveness proposed validated by numerical experimental signals.