摘要
For vibration signals, this paper proposes an ensemble classification method based on stochastic configuration networks (SCNs). Firstly, the time-frequency analysis methods are used to obtain the frequency spectrum signal and time-frequency images. The sample data in the frequency domain and the time-frequency domain can characterize fault information from different perspectives. The hybrid data that consist of the sample data from the two domains are used to build a SCN model. Moreover, a SCNs ensemble method is proposed to solve the fault classification problem, and the sub-classifiers are built to extract fault features from different training data. In the experiment, the bearing and gear fault datasets are used for performance comparison. The experimental results show that the proposed SCNs ensemble model obtains good classification results, and compared with the deep learning methods, the SCN modeling process is more simple and effective for industrial data classification.
学科与关键词
- Industrial data classification using stochastic configuration networks with self-attention learning featuresLi, Weitao; Deng, Yali; Ding, Meishuang; Wang, Dianhui*; Sun, Wei; Li, QiyueNEURAL COMPUTING & APPLICATIONS, 2022, 34(24): 22047-22069.被引 3
- A novel method of rolling bearings fault diagnosis based on singular spectrum decomposition and optimized stochastic configuration networkWang, Shenquan; Lian, Ganggang; Cheng, Chao; Chen, Hongtian*NEUROCOMPUTING, 2024, 574: 127278.被引 6
- Deep subclass reconstruction network for fault diagnosis of rotating machinery under various operating conditionsYu, Hui; Wang, Kai*; Li, Yan; He, MengfanAPPLIED SOFT COMPUTING, 2021, 112: 107755.被引 15
