Research on fault diagnosis model of equipment manufacturing equipment based on deep learning and multi-source sensor data fusion

Aiming at the problem that the fault characteristics of equipment manufacturing equipment are hidden under complex working conditions and the diagnostic accuracy of single sensor data is limited, this paper proposes a fault diagnosis model based on deep learning and multi-source sensor data fusion. Firstly, a multi-source sensing data cube with time-space synchronization is constructed, and heterogeneous data alignment is realized through linear interpolation and Z-score standardization. Then, a parallel CNN-LSTM hybrid architecture is designed to extract the local spatial characteristics and timing dependence of each sensor mode respectively. Finally, the cross-modal attention mechanism is introduced, and the multi-modal features are adaptively weighted to highlight the fault-related sensing information and realize high-precision fault identification. The experiment is based on multi-source data sets of real rotor system (including vibration, temperature, pressure, acoustic emission and other six kinds of sensing data). The results show that the accuracy of the proposed model is 94.7%, and the macro-average F1 score is 0.932, which is significantly better than the traditional methods and single modal models, and it shows good robustness under the conditions of strong noise and class imbalance. The research results provide reliable technical support for intelligent operation and maintenance of equipment manufacturing equipment.