摘要： The paper presents an alternative Vold–Kalman filter order tracking (VKF_OT) method, i.e. adaptive angular-velocity VKF_OT technique, to extract and characterize order components in an adaptive manner for the condition monitoring and fault diagnosis of rotary machinery. The order/spectral waveforms to be tracked can be recursively solved by using Kalman filter based on the one-step state prediction. The paper comprises theoretical derivation of computation scheme, numerical implementation, and parameter investigation. Comparisons of the adaptive VKF_OT scheme with two other ones are performed through processing synthetic signals of designated order components. Processing parameters such as the weighting factor and the correlation matrix of process noise, and data conditions like the sampling frequency, which influence tracking behavior, are explored. The merits such as adaptive processing nature and computation efficiency brought by the proposed scheme are addressed although the computation was performed in off-line conditions. The proposed scheme can simultaneously extract multiple spectral components, and effectively decouple close and crossing orders associated with multi-axial reference rotating speeds. •Adaptive angular-velocity VKFOT proposed as a counterpart of angular-displacement scheme.•Order/spectral waveforms recursively tracked by Kalman filtering based on the one-step state prediction.•Adaptive processing nature characterized and discussed although performed in off-line conditions.•Parameter investigation conducted through processing three synthesized signals.•Comprehensive comparison including computation efficiency among three OT schemes. 出版者： Elsevier Ltd 出版日期： 2016-12-15 出處： Mechanical systems and signal processing, 2016-12, Vol.81, p.148-161 資源來源： Elsevier ScienceDirect Journals Complete 版權： 2016 Elsevier Ltd 識別號： ISSN: 0888-3270 識別號： EISSN: 1096-1216 識別號： DOI: 10.1016/j.ymssp.2016.03.013
