Analysis of the impact of data compression on condition monitoring algorithms for ball screws

Abstract

The overall equipment effectiveness (OEE) is a management ratio to evaluate the added value of machine tools. Unplanned machine downtime reduces the operational availability and therefore, the OEE. Increased machine costs are the consequence. An important cause of unplanned machine downtimes is the total failure of ball screws of the feed axes due to wear. Therefore, monitoring of the condition of ball screws is important. Common concepts rely on high-frequency acceleration sensors from external control systems to detect a change of the condition. For trend and detailed damage analysis, large amounts of data are generated and stored over a long time period (>5 years), resulting in corresponding data storage costs. Additional axes or machine tools increase the data volume further, adding to the total storage costs. To minimize these costs, data compression or source coding has to be applied. To achieve maximum compression ratios, lossy coding algorithms have to be used, which introduce distortion in a signal. In this work, the influence of lossy coding algorithms on a condition monitoring algorithm (CMA) using acceleration signals is investigated. The CMA is based on principal component analysis and uses 17 features such as standard deviation to predict the preload condition of a ball screw. It is shown that bit rate reduction through lossy compression algorithms is possible without affecting the condition monitoring - as long as the compression algorithm is known. In contrast, an unknown compression algorithm reduces the classification accuracy of condition monitoring by about 20 % when coding with a quantizer resolution of 4 bit/sample.