Research on Intelligent Fault Diagnosis Method of Planetary Gearbox Based on ITD and CNN

Accurate fault diagnosis of planetary gearboxes (PGs) is crucial for preventing equipment failures and extending their lifespan. To enhance the efficiency and accuracy of PG fault diagnosis, this paper proposes a novel method integrating intrinsic time scale decomposition (ITD) with convolutional neural network (CNN). The proposed method leverages the demodulation capability of ITD to decompose and reconstruct the original signal, effectively removing irrelevant noise and isolating core fault features, and the learning ability of CNN to accurately identify the fault features. Compared to methods relying on CNN only, the proposed method not only improves the computational efficiency but also achieves higher fault diagnosis accuracy.