Skip to Main Content
Vibration-based machine health monitoring is a well-known technique that is commonly used in machine fault diagnosis applications. In conjunction with existing wireless technologies, even machines that are operated in a hazardous or moveable environment can be monitored wirelessly. However, to facilitate a reliable machine health assessment, various sensory data, particularly the vibration signals, must be acquired at a high sampling rate, generating large packets of data. These requirements are impractical in a wireless data transmission context due to the usual low transmission rate, and lengthy transmission time. One possible solution is to compress the captured numerous sensory data before wireless transmission commences. This paper presents a novel data compression algorithm that combines Empirical Mode Decomposition (EMD) with Differential Pulse Code Modulation (DPCM). EMD is effective in decomposing and identifying any instantaneous changes in non-linear, and non-stationary signals that are caused by anomalous operation of machines. After the data have been decomposed and compressed by EMD, the DPCM is applied to further compress the data through the use of linear predictor and quantizer prior to wireless transmission. A pair of Bluetooth based wireless devices was tailor-made to host the EMD and DPCM for performing wireless communication. The effectiveness of the new algorithm has been verified using real machines, operating in a noisy environment. The results prove that the algorithm provides much faster wireless data transmission by significantly reducing the size of sensory data. Nevertheless, it maintains the integrity of sensory data with negligible reconstruction error at the receiving end of the wireless device.