Bearing mechanical faults classifier based on artificial neural network

Bearing mechanical faults classifier based on artificial neural network

Software

Software implementation of a neural network based on the multiplayer perceptron technique has been created in MATLAB environment. It serves for rolling elements bearing faults classification based on evaluation of the mechanical manifestation. Such quantities (vibration acceleration, ultrasonic and acoustic waves) are measured by appropriate sensors. Neural network has been trained and validated on the real data acquired on the bearing housing for healthy as well as several faulty states of the machine under constant operational conditions. Trained neural network can be easily implemented into microcontroller in the low-performance device, where classification function will be inferred.

Software implementation of a neural network based on the multiplayer perceptron technique has been created in MATLAB environment. It serves for rolling elements bearing faults classification based on evaluation of the mechanical manifestation. Such quantities (vibration acceleration, ultrasonic and acoustic waves) are measured by appropriate sensors. Neural network has been trained and validated on the real data acquired on the bearing housing for healthy as well as several faulty states of the machine under constant operational conditions. Trained neural network can be easily implemented into microcontroller in the low-performance device, where classification function will be inferred.

neural network, multiplayer perceptron, deep learning, supervised learning, bearing failures

neural network, multiplayer perceptron, deep learning, supervised learning, bearing failures

Vysoké učení technické v Brně, CEITEC VUT Laboratoř pokročilých senzorů, B1.04 Purkyňova 656/123 612 00 Brno

K využití výsledku jiným subjektem je vždy nutné nabytí licence

https://ai4di.ceitec.cz/vysledky/ann_bearing_fault_classifier