Abstract
The most commonly used single feature-based anomaly detection method for the complex machinery, such as large wind power equipment, steam turbine generator sets, and reciprocating compressors, exhibits a defect of low-alarm accuracy due to the non-stationary characteristic of the vibration signals. In order to improve the accuracy of fault detection, a novel method based on the Dirichlet process mixture model (DPMM) is proposed. First, the features of the mechanical vibration signals are used to construct the feature space of the equipment. The DPMM modeling method is then applied to self-learn the probabilistic mixture model of the feature space. The normal working condition model is used as the benchmark model. The early fault detection is realized by using a precise difference measurement method based on Kullback-Leibler divergence to calculate the difference between the real-time model and the benchmark model accurately, and by comparing the calculation result with a self-learned alarm threshold. The effectiveness and the adaptability of this novel early fault detection method are verified by comparing it to the single feature-based anomaly detection method and the Gaussian mixture model (GMM)-based early fault detection method.
Highlights
Complex machinery such as large wind power equipment, steam turbine generator sets and reciprocating compressors are widely used in industrial production
The early fault detection is realized by using a precise difference measurement method based on Kullback-Leibler (KL) divergence to accurately calculate the difference between the realtime model and the benchmark model, and by comparing the calculation result with a self-learned alarm threshold
The feature space F = {Fa}ba=1 is constructed by the features of mechanical vibration signals, where b denotes the number of the datasets contained in each training sample, and Fa represents the feature matrix of the ath dataset, which can be defined as f1,1 · · · f1,p fq,1 · · · fq,p q×p where fi,j is the jth feature of the signals collected from the ith measuring point, q is the number of measuring points, and p is the number of features
Summary
Complex machinery such as large wind power equipment, steam turbine generator sets and reciprocating compressors are widely used in industrial production. Due to the Dirichlet process provides a prior information for the distribution of the parameters of the mixture model, DPMM can accurately and automatically determine the number of the components contained in the model according to the observed data [12]. Because of this advantage, DPMM has been widely used in the data mining field for purposes such as text clustering and image segmentation [13]–[16] and has achieved excellent performance. If DPMM can be applied to the statistical distribution self-learning of mechanical vibration signals, the accuracy of complex machinery early fault detection can be significantly improved.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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
