A Benchmark Study for MEMS-Sensors-Based Fault Diagnosis for Rolling Bearings

摘要

Nowadays, high demands are arising for the flexible and low-cost, yet reliable and robust vibration monitoring systems in the context of Industry 4.0 for industrial machines. On the one hand, robust and reliable Condition Monitoring (CM) and fault diagnosis Schemas are expected to detect and identify probable faults even in their early stages and in harsh operating conditions with the lowest possible number of missed detections. On the other hand, however, measuring equipment is required to be low-cost, embedded, and light-weight to be implemented in series production. In this regard, owed to their size and cost, Micro-Electro Mechanical Systems (MEMS)-based sensors receive more attention for CM purposes. The present study aims to investigate capabilities of MEMS accelerometers in vibration-based rolling bearing fault diagnosis, and to benchmark the operation by comparing their results to commercial piezoelectric accelerometers. Therefore, the experimental results of MEMS and commercial piezo sensors are analyzed using state-of-the-art tools in signal processing to highlight usability, advantages, disadvantages, and limitation of using low-cost and low-power MEMS sensors in rolling bearing fault detection applications.