Summary:
Machine learning methods are increasingly used for rotating machinery monitoring. Usually at system set up, only data of the machinery in healthy conditions, the so-called nominal data, are available for the machine learning phase. This type of training data enables fault detection capabilities and several methods such as Gaussian Mixture Model, One Class Support Vector Machines and Auto Associative Neural Networks (Autoencoders) have been already proved successful for this task. However, in some predictive maintenance applications, information on the type of defect may represent a key element for producing actionable information, e.g. to reduce diagnostic burden and optimize spare procurement. This requires to define classification strategies based on machine learning even in absence of data representing the behaviour of the system with defects. In this study we present an approach that uses only nominal vibration data to train an autoencoder which will enable at same time fault identification and fault classification tasks. As faulty data are expected to possess information content which is structured differently from the healthy ones their reconstruction at output will result inaccurate. In conventional anomaly detection approaches, the module of the reconstruction error, defined as the difference between output and input, is uses to determine an unusual input such as faults. The proposed approach represents a step forward as here a single autoencoder is used both for detection and classification. The underlying idea is that the components of the reconstruction error vector whose module is used to trigger fault identification in classical autoencoder approaches contain the information of the fault type. This way the analysis of the different components of the reconstruction error allows to differentiate the different types of faults. Two methods to analyse the components of the reconstruction error vector will be discussed and their respective test results will be presented Test data have been generated with a machine fault simulator to produce 3 different types of bearing defects with different load, speed and noise conditions. A dataset of about 10000 vibration signals has been used to evaluate the classification algorithms and to benchmark them with a supervised approach. The results obtained using the autoencoder method do not achieve the same performances as the conventional supervised learning algorithms. However, they proved to be 88% accurate in classification when SNR is above 0dB with the ranking based method overperforming the barycentre one.
SWITCH edu-ID
Administration
