Summary: | 碩士 === 國立中央大學 === 機械工程研究所 === 97 === The study presents and implements one advanced VKF_OT approaches, i.e. adaptive angular-velocity VKF_OT techniques, to overcome the deficiencies of original VKF_OT schemes for condition monitoring and diagnosis of rotary machinery. This thesis comprises the theoretical derivation, numerical implementation, and experimental justification. Comparisons of the improved VKF_OT schemes to the original ones are accomplished through processing three synthetic signals composed of different order/spectral components. Some parameters such as the weighting factor, the correlation matrix of process noise, and unwanted signatures/noise, which influence tracking performance, are investigated in the study. Additionally, three experimentations are designed to justify the effectiveness of applying the developed VKF_OT techniques in practical works. These proposed schemes can simultaneously extract multiple order/spectral components, and effectively decouple close or crossing orders associated with multi-axial reference rotating speeds. Furthermore, the adaptive angular-velocity VKF_OT scheme based on the Kalman filter can be calculated on-line and implemented as a real-time processing application.
|