Deep networks based mechanical intelligent diagnosis has been recently attracting considerable attentions with the development of Industry 4.0. Unfortunately, a more practical diagnostic scenario, i.e. unsupervised partial transfer diagnosis (UPTD), has not yet been well addressed. In view of this, a novel unsupervised intelligent diagnosis framework named Classifier Inconsistency based Domain Adaptation network (CIDA) is proposed in this work. In this approach, two discriminative one-dimensional convolutional networks are designed as the basic architecture. The source samples of the same categories as the target domain are then identified and emphasized to boost positive network training. Meanwhile, the classifier inconsistency is introduced to guide the model to learn discriminative and domain-invariant representations for the correct classification of unlabeled target data. Extensive experiments on two datasets are conducted to evaluate the proposed method. Additionally, five popular methods are selected for comparison. The comprehensive results validate the effectiveness and superiority of the proposed approach.

Machine learning models have been widely successful in the field of intelligent fault diagnosis. Most of the existing machine learning models are deployed in static environments and rely on precollected datasets for offline training, which makes it impossible to update the models further once they are established. However, in the open and dynamic environment in reality, there is always incoming data in the form of streams, including new categories of data that are constantly generated over time. In addition, the operating conditions of mechanical equipment are time-varying, which results in continuous stream data that are nonindependently and homogeneously distributed. In industrial applications, the diagnosis problem of nonindependent and identically distributed continuous streaming data is referred to as the cross-domain class incremental diagnosis problem. To address the cross-domain class incremental problem, a novel cross-domain class incremental broad network (CDCIBN) is proposed. Specifically, to solve the nonindependent identically distributed problem, a novel domain-adaptation learning loss function is first designed, which enables the conventional broad network to handle the category increment task well. Then, a cross-domain class incremental learning mechanism is designed, which learns new categories while retaining the knowledge of old categories well enough without replaying old category data. The effectiveness of the proposed method is evaluated through multiple mechanical failure increment cases. Experimental analysis demonstrates that the designed CDCIBN has significant advantages in the variable working condition class incremental application.

Envelope demodulation based on vibration data is widely used for the fault detection of rolling element bearing, which yet largely relies on the high signal to noise ratio of signal. In practical scenarios, because of the existence of various interfering components, it is necessary to estimate the fault-sensitive frequency band for feature enhancement. However, many of current approaches are likely to produce misleading results for that they are robust only for part of these interferences. Facing with this problem, an alternative method, named as ALKurtogram, is proposed by aligning the frequency division strategy of traditional fast kurtogram (FK) with a new indicator, the averaged local kurtosis (ALK). ALK measures both the local and overall impulsiveness of objective component, which intrinsically makes up for some drawbacks of previous methods under multi-interference condition for fault feature extraction. The inherent nature of ALK also allows it to be compatible with the time-varying speed, hence more widespread industrial applications can be achieved. Furthermore, a harmonic energy index (HEI) is defined by following the result of ALKurtogram to ascertain the order of principle component for bearing status recognition. The effectiveness and superiority of proposed method are validated experimentally by comparing with FK, the results prove that it is a powerful tool for reliable monitoring of bearing.