Classification dynamic strain patterns in switching devices measured with Fiber Bragg Grating

Vibration measurement systems are developed to assess the operational condition of electrical machines and electromechanical devices in different types of applications. This work presents a measurement system with Fiber Bragg Grating. The experimental setup was developed for the analysis of vibration signals from small size electromechanical switching devices. With statistical criteria, the selection of features extracted from measured signals, such as the switching time and Power Spectral Density, presents 98 % accuracy in classifying four distinct operational conditions using Support Vector Machine algorithm. The results validated the proposed measurement system, which allows the possibility to extend the application to more complex devices in future studies.