Fault Detection and Severity Level Identification of Spiral Bevel Gears under Different Operating Conditions Using Artificial Intelligence Techniques

اسم المؤلف

Syed Muhammad Tayyab, Steven Chatterton and Paolo Pennacchi

التاريخ

15 سبتمبر 2022

التصنيف

هندسة الإنتاج والتصميم الميكانيكى

المشاهدات

320

التقييم

Loading...

التحميل

اضغط هنا للتحميل

بحث بعنوان
Fault Detection and Severity Level Identification of Spiral Bevel Gears under Different Operating Conditions Using Artificial Intelligence Techniques
Syed Muhammad Tayyab, Steven Chatterton * and Paolo Pennacchi
Department of Mechanical Engineering, Politecnico di Milano, Via G. La Masa 1, 20156 Milan, Italy;
[email protected] (S.M.T.); [email protected] (P.P.)

• Correspondence: [email protected]; Tel.: +39-02-2399-8442
  Abstract: Spiral bevel gears are known for their smooth operation and high load carrying capability;
  therefore, they are an important part of many transmission systems that are designed for high
  speed and high load applications. Due to high contact ratio and complex vibration signal, their
  fault detection is really challenging even in the case of serious defects. Therefore, spiral bevel gears
  have rarely been used as benchmarking for gears’ fault diagnosis. In this research study, Artificial
  Intelligence (AI) techniques have been used for fault detection and fault severity level identification
  of spiral bevel gears under different operating conditions. Although AI techniques have gained much
  success in this field, it is mostly assumed that the operating conditions under which the trained AI
  model is deployed for fault diagnosis are same compared to those under which the AI model was
  trained. If they differ, the performance of AI model may degrade significantly. In order to overcome
  this limitation, in this research study, an effort has been made to find few robust features that
  show minimal change due to changing operating conditions; however, they are fault discriminating.
  Artificial neural network (ANN) and K-nearest neighbors (KNN) are used as classifiers and both
  models are trained and tested by using the selected robust features for fault detection and severity
  assessment of spiral bevel gears under different operating conditions. A performance comparison
  between both classifiers is also carried out.
  Keywords: fault detection; fault severity level identification; artificial intelligence (AI); artificial
  neural network (ANN); K-nearest neighbors (KNN); features extraction
  Conclusions
  In this study, fault detection and severity level identification of spiral bevel gears are
  carried out under different operating conditions by using two AI models, ANN and KNN,
  as classifiers. Time domain statistical features were extracted from the vibration data of
  spiral bevel gears, one with normal health condition and two with faulty conditions at
  different severity levels, in order to train the classifiers. The performance of both classifiers
  in terms of fault detection and severity level identification accuracy gradually degraded as
  the operating conditions under which the models were deployed for predictions deviated
  farther away from the operating conditions under which the models were trained. The
  performance degradation was due to the higher sensitivity of most of the features underconsideration towards the operating conditions. Variation in most of the features due to
  operating conditions was much more prominent than compared to their change because
  of the fault or fault severity level. Therefore, most of the features were misleading the
  classifiers. The features were found more sensitive to change in speed than compared to
  change in load. Three features (rms, Energy-I and Energy-II) were identified as robust features which showed least sensitivity to operating conditions but were fault discriminative
  and demonstrated an increasing trend with respect to fault severity level. ANN and KNN
  performed predictions with 100% accuracy under all operating conditions while using
  only robust features. Thus, the performance of ANN and KNN classifiers was significantly
  improved for fault detection and severity level identification of spiral bevel gears under
  different operating conditions by eliminating misleading features, which were sensitive to
  operating conditions, and selecting the robust features that are less sensitive to operating
  conditions but were also fault discriminative. The overall performance of ANN and KNN
  classifiers was found almost comparable to one another.

كلمة سر فك الضغط : books-world.net
The Unzip Password : books-world.net