A LSTM Neural Network Approach using Vibration Signals for Classifying Faults in a Gearbox

Mathematics

• Vibration Signal 100%
• Gearbox 90%
• Memory Term 68%
• Fault 60%
• Neural Networks 48%
• Feature Space 47%
• Cross-validation 38%
• Learning 31%
• Intrinsic Mode Function 29%
• Empirical Mode Decomposition 28%
• Voting Systems 27%
• Multi-class 22%
• Decomposition Method 18%
• Second derivative 17%
• Fold 15%
• Model 5%

Physics & Astronomy

• transmissions (machine elements) 79%
• classifying 58%
• vibration 34%
• learning 28%
• voting 20%
• decomposition 9%

Engineering & Materials Science

• Long short-term memory 60%
• Deep learning 46%
• Neural networks 31%
• Derivatives 21%
• Decomposition 18%