Sistema inteligente de diagnóstico de fallas en máquinas rotativas usando el enfoque de aprendizaje automático

Spindle failures diagnosis in high-speed machining centers is critical in manufacturing systems, since early detection can save a representative amount of time and cost. The fault diagnosis systems usually have two blocks: feature extraction and classification, the feature extraction affects the performance of prediction model, and the essential information is realized by identification of abstract and representative high-level features. Deep Learning (DL) provides an effective way to extract the features of raw data, without prior knowledge compared with traditional Machine Learning (ML) methods. A feature learning approach was applied using 1D CNN that works directly with raw vibration signals. The network structure consists of small convolutional kernels to realize a nonlinear mapping and extract features, the classifier is a Softmax layer. The method has achieved a satisfactory performance in terms of prediction accuracy reaching an ∼99% using three bearing databases, the processing time is suitable for real-time applications with ∼8ms per signal, the repeatability has a low standard deviation ∼0.25% and achieves an acceptable network generalization ability.