A hybrid optimized approaches for ball bearing state prognosis for effective decision making

Manufacturing Industries (MI) has developed in the past years, by including different types of machines such as the Rotating Machines (RM) to deliver High-Quality Products (HQP), however those machines are prone to failure, which effects on the production process and leading to high economic loses, the purpose behind this study is to present a new strategy that support’s the decision-making to avoid any unexpected breakdowns in the MI. A novel hybrid technique that combines the Adaptive Neuro Fuzzy Inference System (ANFIS), the Hybrid Whale Grey Wolf Optimizer (HWGO), and the state space (SS) approach using the Vibration Condition Monitoring (VCM) signals to determine the Remaining Useful Life (RUL) is introduced. The proposed approach was trained using features selected by the Decision Tree (DT) algorithm. A comparative analysis is conducted against the HWGO with 12 metaheuristic algorithms in terms of optimization. The SS model was applied for forecasting the future RUL values depending on the measured values by the ANFIS-HGWO. The overall results confirm the outperforming of the ANFIS-HWGO-SS compared to different contributions available in the literature using the PRONOSTIA database in terms of performance measure, Therefore, the ANFIS-HWGO-SS approach is a reliable tool for determining the RUL before any unexpected breakdown and further supports the decision-making by offering a vital timeframe for making the proper action before a failure occurs which can be applied for other machine-related tasks to ensure stability by decreasing the failure of the machine, reliability of the RM by providing HQP, security by avoiding accidents in MI, marking a significant step towards enhanced operational efficiency and sustainability. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.