A multi-sensor fault diagnosis method for rotating machinery based on improved fuzzy support fusion and self-normalized spatio-temporal network

Recently, the fault diagnosis of rotating machinery based on deep learning has achieved increasingly widespread applications. However, it is often difficult to achieve the expected results by relying on a single sensor due to the limited information obtained by the single sensor and the susceptibility to the influence of the additive noise. To address the above problems, this paper proposes a multi-sensor fusion fault diagnosis method for rotating machinery based on improved fuzzy support fusion and self-normalized spatio-temporal network to enhance feature learning while achieving multi-sensor data fusion. This method includes a data pre-processing module, a fusion module and a fault recognition module. In the first module, a complete ensemble empirical mode decomposition with adaptive noise algorithm is introduced to decompose and reconstruct the multi-source sensor signals, thereby reducing the impact of environmental noise on data quality. In the fusion module, a data fusion algorithm based on improved fuzzy support is designed to achieve the data-level fusion of multi-source sensors. By introducing the self-normalized properties into the convolutional structure with bi-directional gated recurrent unit, a self-normalized spatio-temporal network is designed in the fault recognition module to perform the fault diagnosis of rotating machinery. The experimental results show that the proposed method can achieve high quality data-level fusion and outperforms the state-of-the-art fault diagnosis methods in terms of fault classification.