Paper Title
Merging Artificial Intelligence with Digital Twins for Fault Prediction and Classification Using Suspension’s Primary Springs
Abstract
Research on Artificial Intelligence (AI) based equipment failure diagnoses is significant; However, integrating AI
with Digital Twins (DTs) remains challenging. The computational methods could not be fault-free due to the complexity of
faults and the interaction between a failure and system response for fault positives and negatives. A lack of intelligent
management, monitoring, and feedback constitutes a lack of intelligent supervision. An Advanced Digital Models (ADTM)
framework with an Artificial Neural Network (ANN) is presented. Combining the ADTM framework with ML through
ANN and Support Vector Machines (SVM) tackles these current challenges and overcomes systems‘ complex issues. An
investigation into a suspension system‘s case study demonstrates the viability of ADTMs. For the first time, integration
between DTs and ML for monitoring the system‘s current and futuristic conditions is fully presented with a prediction error
of less than 0.0001 %. The ANN model was improved, classifying the system‘s faults with regression accuracy throughout
training, validation, and testing were 0.99997, 0.99954, and 0.99931, respectively, for a total accuracy of 0.99986.
Keywords - Digital Twin, Hybrid Twin, IoT Healthcare, and Machine Learning