一种基于人工智能的机器人核心零部件故障诊断方法
By combining short-time Fourier transform and Chebyshev graph convolutional network with reinforcement learning, the noise interference problem of vibration signal of harmonic reducer was solved, and efficient fault feature extraction and accurate fault diagnosis were achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NORTHEASTERN UNIV FOSHAN GRADUATE SCHOOL OF INNOVATION
- Filing Date
- 2024-07-17
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional methods for diagnosing rotating machinery faults struggle to effectively extract fault features when processing non-stationary vibration signals. In particular, the vibration signals from harmonic reducers are characterized by complex noise, making signal processing difficult. Furthermore, graph convolutional networks are prone to over-smoothing during training, which affects accuracy.
The topological features of the vibration signal of the harmonic reducer are extracted by short-time Fourier transform. Combined with Chebyshev graph convolutional network and reinforcement learning method, the network model is trained with graph structure data and the network layers are adaptively adjusted to achieve fault diagnosis.
It effectively suppresses noise interference, improves the accuracy of fault diagnosis, and ensures the reliability and precision of test results.
Smart Images
Figure CN119089337B_ABST