Fault Diagnosis Method of Mechanical Equipment Based on Convolutional Neural Network

Abstract

Mechanical equipment is becoming much larger, more precise and more autonomous in current industrial society. The mechanical equipment fault detection is entering the age of 'big data' for much more monitoring points and sampling rate. Traditional diagnosis methods based on "signal processing feature extraction + machine learning classification" require a large amount of signal processing technology and diagnostic experience and can no longer meet the requirements of mechanical 'big data'. To solve this problem, an important part bearing in mechanical equipment is taken as the research object, and a diagnosis method based on convolutional neural network is proposed. This method uses the vibration signal as the monitoring signal and uses the Fourier transform to generate the vibration signal spectrum picture as the input of the whole system. Using the powerful feature extraction capability of convolutional neural network can automatically complete fault feature extraction and fault identification. The results show that the proposed method is able to not only adaptively mine available fault characteristics from the data, but also obtain higher identification accuracy than the existing methods.