Optimal compression of vibration data with lifting wavelet transform and context-based arithmetic coding

Abstract

Abstract-This paper proposes an adaptive vibration signal compression scheme composed of a lifting discrete wavelet transform (LDWT) with set-partitioning embedded blocks (SPECK) that efficiently sorts the wavelet coefficients by significance. The output of the SPECK module is input to an optimized context-based arithmetic coder that generates the compressed bitstream. The algorithm is deployed as part of a reliable and effective health monitoring technology for machines and civil constructions (e.g. power generation system). This application area relies on the collection of large quantities of high quality vibration sensor data that needs to be compressed before storing and transmission. Experimental results indicate that the proposed method outperforms state-of-the-art coders, while retaining the characteristics in the compressed vibration signals to ensure accurate event analysis. For the same quality level, up to 59.41% bitrate reduction is achieved by the proposed method compared to JPEG2000.

Adaptive event-triggered anomaly detection in compressed vibration data

“…Nowadays, industrial CM systems use multiple sensors and generate continuous streams of raw noise and vibration data. Compression can be used to reduce storage requirements and data transfer bandwidth to the control room outlined in the authors' previous work [2]. The effects of compression on the trigger mechanism have not been previously investigated although it could have negative effects on the accuracy and robustness of the subsequent thresholding and decision making processes.…”

“…The benefit of the VSC algorithm is that a single coded bitstream can be used to decode the signal at any rate less than or equal to the coded rate, to provide the best reconstruction quality [2]. The algorithm is very useful for remote CM systems that can perform an initial base layer decoding so the AETM can quickly inspect the vibration signal and decide if it should be fully decoded (adding more detailed information to the base layer) for further analysis, or it is good enough to be categorized as normal health status signal.…”

“…However, the understanding of the failures resulting from transient events in retrospect becomes more difficult if the reduced sampling frequency means that valuable information has been lost. Another approach is to apply data compression algorithms [2] that can adapt its acquisition conditions and compression ratio without loss of critical information.…”

“…In this paper the vibration data is collected from multiple sensors and compressed using the techniques initially presented in [2] in real-time using a low precision base layer and high precision refinement layer. An Adaptive Event Trigger Mechanism (AETM) system is optimized to process only the base layer so no high precision decoding is required under nominal machine operation.…”

“…The diagnostic model performs the on-line analysis and quantification of the machine health status. The constructed diagnostic model adapts the trigger condition dynamically guided by data changes at the diagnostic level during machine operation.The VSC algorithm proposed in[2] is part of a reliable and effective health monitoring technology for machines and civil infrastructure (e.g. power generation system), as shown inFig.…”