Anomaly Detection in Rotary Equipment using hybrid model with HGSO Optimization

The reliability of rotating machinery is critical for continuous industrial operations in the context of Industry 4.0. The essential assets of industries are motors, pumps, fans, generators, and compressors, these equipment's have enormous importance, since their failure may lead to significant production losses and decreased competitiveness. The need for machine health monitoring is very essential. This paper proposes a hybrid Support Vector Machine (SVM) and Extreme Gradient Boosting (XGBoost) predictive maintenance model for anomaly detection. The model is optimized using a bio-inspired algorithm Henry Gas Solubility Optimization (HGSO) to attain the optimal hyper parameters and accuracy. The proposed model uses ISO 10816 vibration diagnostic standards to evaluate machine severity based on velocity and vibration data. The results of the proposed optimized model provide 94% accuracy. The model helps maintenance operators to analyze health severity status and avoid unplanned machine shutdowns