Abstract
The air compressor is one of the desired mechanical equipment used for producing compressed air, which is utilized for performing various industrial and domestic functions. Its operation involves several rotating and fluctuating members which fail due to several miscellaneous reasons as the members prone to dynamic working environment quite frequently. The deficiencies create huge impact over the overall performance and thus leads to economic losses associated with system seizure. It is now essential to predict the occurrence of faults at earlier stages in order to avoid major shutdowns. Hence, in this article, a data modelling study using a machine learning algorithm is proposed. Initially, the vibration signals are measured as physical parameters from the compressor test rig as it contains critical information regarding the system working conditions instantly. The statistical features were extracted from the acquired signals and by using the J48 algorithm the most prominent features were selected. These selected features were classified using Multilayer Perceptron and its performance in fault classification was presented
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call