Loading...

Intelligent Fault Diagnosis using Multiple Sensor Data Fusion for Detecting Misalignment and Unbalance

Yadegari, Mohammad Erfan | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57560 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Behzad, Mehdi; Arghand, Hesam Al-Din
  7. Abstract:
  8. Intelligent predictive maintenance is recognized as a cornerstone of Industry 4.0, where intelligent software is employed for the early detection of faults and the prevention of unexpected failures. Recent research indicates that the integration of multi-sensor data for fault diagnosis of gearboxes and bearings, using artificial intelligence models, has been successful. However, conventional methods face several challenges. These include an over-reliance on the signal characteristics of a single sensor and the impracticality of applying intelligent learning methods, particularly deep learning, despite their high potential, due to the unavailability of sufficiently large and diverse industrial datasets within the country. In this study, an algorithm was developed for multi-sensor data fusion through the use of combined features and decision-level information fusion utilizing fuzzy inference engines. This algorithm, by processing data from sensors installed at various points on equipment, is capable of detecting equipment faults. Initially, the necessary features and rules for fault detection were gathered and compiled in consultation with experts and through the review of standards. Subsequently, the decision tree intelligent algorithm was implemented to determine appropriate threshold values and enhance the generalizability of the rules. Finally, two separate fuzzy systems were developed for diagnosing each fault, and their accuracy was evaluated. Additionally, the performance of the proposed algorithm was compared with common machine learning models. The results demonstrate that the developed algorithm can accurately detect misalignment and imbalance faults in rotary equipment
  9. Keywords:
  10. Intelligent Algorithms ; Fault Diagnosis ; Multi-Sensor Data Fusion Network ; Fuzzy Inference System ; Misalignment Fault ; Imbalanced Data ; Preventive Maintenance

 Digital Object List

 Bookmark

...see more