Research on a feature extraction method for local faults in planetary gearboxes based on improved dynamic time warping

Aiming at the local fault diagnosis of planetary gearbox gears, a feature extraction method based on improved dynamic time warping (IDTW) is proposed. As a calibration matching algorithm, the dynamic time warping method can detect the differences between a set of time-domain signals. This paper applies the method to fault diagnosis. The method is simpler and more intuitive than feature extraction methods in the frequency domain and the time-frequency domain, avoiding their limitations and disadvantages. Due to the shortcomings of complex calculation, singularity and poor robustness, the paper proposes an improved method. Finally, the method is verified by envelope spectral feature analysis and the local fault diagnosis of gears is realised.

Download Full-text

Time-Domain Fault Diagnosis Method of Mechanical and Electrical Equipment Based Improved Dynamic Time Wraping

Key Engineering Materials ◽

10.4028/www.scientific.net/kem.693.1539 ◽

2016 ◽

Vol 693 ◽

pp. 1539-1544 ◽

Cited By ~ 1

Author(s):

Zhi Wu Shang ◽

Zhen Wu Liu ◽

Ya Feng Li ◽

Tai Yong Wang

Keyword(s):

Feature Extraction ◽

Fault Diagnosis ◽

Time Domain ◽

Dynamic Time Warping ◽

Statistical Parameter ◽

Electrical Equipment ◽

Time Warping ◽

Trend Prediction ◽

Signal Characteristics ◽

Dynamic Time

Dynamic time warping used in speech recognition widely was migrated to fault feature extraction and diagnosis in time domain. Integration of phase compensation, slope weighted, derivative, sliding window connection, fast dynamic time planning method is applied to dynamic time warping method. And a new method of time-domain signal feature extraction and fault diagnostic based on improved dynamic time warping method of mechanical and electrical equipment was proposed. Identification and localization of fault signal characteristics may be done by improving dynamic time warping method to obtain a residual signal sequences with fault characterized sidebands and selecting the statistical characteristic parameters such as peak, RMS, kurtosis spectrum to complete identification and localization of fault signal characteristics. New time-domain fault trend prediction method of mechanical and electrical equipment was established based on new statistical parameter Thikat. A new idea and target was provided for fault diagnosis of mechanical and electrical equipment.

Download Full-text

A Novel Local Time-Frequency Domain Feature Extraction Method for Tool Condition Monitoring Using S-Transform and Genetic Algorithm

IFAC Proceedings Volumes ◽

10.3182/20140824-6-za-1003.02515 ◽

2014 ◽

Vol 47 (3) ◽

pp. 3516-3521 ◽

Cited By ~ 4

Author(s):

Javad Soltani Rad ◽

Youmin Zhang ◽

Chevy Chen

Keyword(s):

Genetic Algorithm ◽

Feature Extraction ◽

Frequency Domain ◽

Condition Monitoring ◽

Extraction Method ◽

Tool Condition Monitoring ◽

Feature Extraction Method ◽

Time Frequency ◽

Tool Condition ◽

S Transform

Download Full-text

Mechanical Fault Diagnosis Using Color Image Recognition of Vibration Spectrogram Based on Quaternion Invariable Moment

Mathematical Problems in Engineering ◽

10.1155/2015/702760 ◽

2015 ◽

Vol 2015 ◽

pp. 1-11 ◽

Cited By ~ 4

Author(s):

Liang Hua ◽

Yujian Qiang ◽

Juping Gu ◽

Ling Chen ◽

Xinsong Zhang ◽

...

Keyword(s):

Neural Network ◽

Feature Extraction ◽

Fault Diagnosis ◽

Extraction Method ◽

Learning Algorithm ◽

Probabilistic Neural Network ◽

Invariant Moment ◽

Feature Extraction Method ◽

Time Frequency ◽

Neural Network Algorithm

Automatic extraction of time-frequency spectral image of mechanical faults can be achieved and faults can be identified consequently when rotating machinery spectral image processing technology is applied to fault diagnosis, which is an advantage. Acquired mechanical vibration signals can be converted into color time-frequency spectrum images by the processing of pseudo Wigner-Ville distribution. Then a feature extraction method based on quaternion invariant moment was proposed, combining image processing technology and multiweight neural network technology. The paper adopted quaternion invariant moment feature extraction method and gray level-gradient cooccurrence matrix feature extraction method and combined them with geometric learning algorithm and probabilistic neural network algorithm, respectively, and compared the recognition rates of rolling bearing faults. The experimental results show that the recognition rates of quaternion invariant moment are higher than gray level-gradient cooccurrence matrix in the same recognition method. The recognition rates of geometric learning algorithm are higher than probabilistic neural network algorithm in the same feature extraction method. So the method based on quaternion invariant moment geometric learning and multiweight neural network is superior. What is more, this algorithm has preferable generalization performance under the condition of fewer samples, and it has practical value and acceptation on the field of fault diagnosis for rotating machinery as well.

Download Full-text