Micro-electromechanical systems sensor characterization for acoustic prognostic health management

Abstract

Prognostic Health Management (PHM) is the process used to optimize machinery use by detecting faults and predicting failures and a machine’s remaining useful life. PHM is especially important to electro-mechanical systems in order to maximize a system’s availability and effective operation. The use of Micro-Electromechanical Systems (MEMS) accelerometers and microphones for PHM analysis is beneficial because it enables non-destructive acoustic testing with reduced cost, reduced size, and more efficient vehicle integration with other electronics than current methods. This research quantitatively studies a variety of MEMS sensors to understand how they measure acoustic data in a lab setting and on a truck with a diesel engine compared to piezoelectric accelerometers and condenser microphones. Diesel engines are common in many vehicles and hybrid-electric power systems, so the results will be applicable to a variety of situations. By applying signal processing techniques and PHM condition indicators to the data, standardized metrics can be developed to determine the best sensor for different PHM analyses.