An Intelligent Optimized Digital Twins Framework for Fault Diagnosis in Complex Control Systems | ||||
| Journal of Integrated Engineering and Technology | ||||
| Article 5, Volume 1, Issue 1, 2024, Page 55-68 PDF (1000.62 K) | ||||
| Document Type: Review Article | ||||
| DOI: 10.21608/jiet.2024.274041.1005 | ||||
 View on SCiNiTO | ||||
| Author | ||||
Samar M zayed 
| ||||
| 25 Agricultural street, Cairo, Egypt | ||||
| Abstract | ||||
| Digital Twins (DT) is considered as the backbone of several industrial systems in manufacturing category. The DT strategy has a vital role for dataset generation especially in fault prediction and diagnosis aspects. Recently, these approaches are considered the tending in research by utilizing the support of Artificial Intelligence (AI) techniques for critical industrial applications. The virtual assets of DT can produce a performance that is close to the real counterpart, which is an opportunity for fault diagnosis and prediction under different fault conditions. Therefore, this study proposes an intelligent AI-based framework that is based on Genetic Algorithm (GA) and machine learning classifiers (MLCs) such as Logistic Regression (LR), Linear Discriminant Analysis (LDA), Naïve Bayes (NB), and K-nearest neighbors (KNN) for industrial digital twins systems namely Transmission System (TS) model. The proposed hybrid GA–ML framework is validated using a simulated dataset which is generated from TS model. The proposed framework achieves superior results for MLCs such as LR, LDA, NB, and KNN with accuracy equal to 96.5%, 98.3%, 97.4%, and 97.4 % compared with the ordinary MLCs with 87.3%, 87.3%, 82.5, and 85.7% respectively. Also, it is considered as a superior compared with the existing model’s performance for diagnosing the complex future faults. So, the proposed framework will efficiently help for diagnosing and detecting faults in several manufacturing inspects. | ||||
| Keywords | ||||
| Keywords: Digital Twins(DT); Genetic Algorithm(GA); Machine Learning(ML); Fault Diagnosis; Industrial Control Systems | ||||
|
Statistics Article View: 72 PDF Download: 78 |
||||
