Structural Damage Identification of Composite Rotors Based on Fully Connected Neural Networks and Convolutional Neural Networks

Low-Cost Wireless Structural Health Monitoring of Bridges

Nowadays, low-cost accelerometers are getting more attention from civil engineers to make Structural Health Monitoring (SHM) applications affordable and applicable to a broader range of structures. The present accelerometers based on Arduino or Raspberry Pi technologies in the literature share some of the following drawbacks: (1) high Noise Density (ND), (2) low sampling frequency, (3) not having the Internet's timestamp with microsecond resolution, (4) not being used in experimental eigenfrequency analysis of a flexible and a less-flexible bridge, and (5) synchronization issues. To solve these problems, a new low-cost triaxial accelerometer based on Arduino technology is presented in this work (Low-cost Adaptable Reliable Accelerometer-LARA). Laboratory test results show that LARA has a ND of 51 µg/√Hz, and a frequency sampling speed of 333 Hz. In addition, LARA has been applied to the eigenfrequency analysis of a short-span footbridge and its results are compared with those of a high-precision commercial sensor.

Self-Supervised Joint Learning Fault Diagnosis Method Based on Three-Channel Vibration Images

The accuracy of bearing fault diagnosis is of great significance for the reliable operation of rotating machinery. In recent years, increasing attention has been paid to intelligent fault diagnosis techniques based on deep learning. However, most of these methods are based on supervised learning with a large amount of labeled data, which is a challenge for industrial applications. To reduce the dependence on labeled data, a self-supervised joint learning (SSJL) fault diagnosis method based on three-channel vibration images is proposed. The method combines self-supervised learning with supervised learning, makes full use of unlabeled data to learn fault features, and further improves the feature recognition rate by transforming the data into three-channel vibration images. The validity of the method was verified using two typical data sets from a motor bearing. Experimental results show that this method has higher diagnostic accuracy for small quantities of labeled data and is superior to the existing methods.