A Comprehensive Review on Real-Time Performance Assessment & Operational Fault Monitoring in Hydromachinery

Optimization and Improvement of Operation and Maintenance Efficiency of Condition Monitoring Technology for Offshore Wind Turbines

This study proposes an integrated condition-monitoring and predictive-maintenance framework for offshore wind turbines operating in harsh marine environments. To address the challenges of signal degradation, environmental interference, and limited fault-warning capability, a multi-source sensing architecture is developed based on risk-driven sensor deployment, edge-side signal fusion, and intelligent health assessment. Vibration, temperature, strain, and operational signals are adaptively processed through variance-weighted fusion and denoising strategies to improve data reliability. A CNN– LSTM hybrid model is employed for fault feature extraction and temporal degradation analysis, while a digital-twin-driven health assessment framework is used to quantify health indices and remaining useful life. Maintenance scheduling is further optimized by integrating equipment health conditions, resource constraints, and operational windows. Field validation in an offshore wind farm demonstrates that the proposed diagnostic model achieves a fault identification accuracy of 96.2%, while the predicted remaining useful life of the main bearing decreases from 180 days to 16 days before failure. The proposed framework establishes a closed-loop process linking signal acquisition, intelligent diagnosis, lifetime prediction, and maintenance decision-making, providing an effective engineering solution for reliable condition monitoring and intelligent operation of offshore energy systems.

Recent developments in data-driven and hybrid intelligent techniques for hydropower fault analysis

This review summarizes recent developments in machine learning, deep learning, physics-informed neural networks, and digital twin-assisted approaches for hydropower fault analysis, providing a possible way to integrate real-time monitoring, fault diagnosis, and operational assessment within a unified platform.

Vibration-induced failures in industrial ventilation systems: A predictive maintenance approach

Advanced Bearing Condition Monitoring for Energy Production Machinery via Koopman Dynamics

Monitoring the health of bearings in industrial rotating machines is a major challenge for ensuring the reliability and continuous operation of installations. Conventional fault detection methods, based on multivariate control charts such as Hotelling’s T2, multivariate exponentially weighted moving average, or multivariate cumulative sum control chart, are limited by the complex nonlinear dynamics of the system. In this article, we propose an innovative monitoring approach based on the Koopman operator, allowing the linearization of a nonlinear system in an observed space and the application of drift detection techniques via an extended T2 control chart. The study is based on two experimental approaches: one using controlled simulated data to analyze the responsiveness and robustness of the model, and the other applied to real data from an industrial turbogenerator monitoring the vibrations, temperatures, and speeds of the front and rear bearings. Comparative results show that the Koopman-based T2 map detects defects earlier, with better accuracy under noise and a reduced false alarm rate compared to conventional methods. The integration of wavelet preprocessing, statistical feature extraction by sliding windows, and PCA representation of the trajectories enhances the robustness and interpretability of the model.

Smart Fault Diagnosis in Turbine Bearings: Using Control Charts for Predictive Maintenance and Machine Health Monitoring

Proactive detection of faults in turbine bearings is key to ensuring system reliability in industrial systems. This work introduces the use of Moving Average (MA) and Exponentially Weighted Moving Average (EWMA) control charts for diagnosing bearing faults in the context of predictive maintenance. Unlike other statistical control charts, MA and EWMA control charts provide an advantage in that they employ a dynamic method of identifying trends over time through smoothing out variations as well as quickening the detection of gradual trends in system performance. MA charts are well-suited to find mean behavioral trends since they offer a strong perspective of medium-term behavioral changes. EWMA diagrams show better accuracy in spotting minor, slow variations, which is especially helpful for early-stage fault detection in high-sensitivity settings like turbine bearings. Implementation of the proposed method for real-world turbine operating conditions is shown to demonstrate the potential of using MA and EWMA control charts in monitoring vibration as well as speed anomalies prior to and after maintenance. Results are presented as proof of efficacy in identifying faults, aiding in decision-making on maintenance, and improving the lifespan and system operability of turbomachinery.

Operational Availability Assessment of Tidal Stream Turbines Using Environmental Data and Fuzzy Logic Inference

In hybrid energy systems, maintaining an optimal scheduling strategy for real-time distribution systems, particularly in triple hybrid power generation units, remains a critical challenge. The lack of an efficient real-time observability platform for off-grid hybrid units directly impacts scheduling priorities. In this work, a novel operational condition monitor that has a data-driven predictive mechanism for determining the instant states of each tidal stream turbine is proposed. Environmental variables are first preprocessed using a multivariate fuzzy logic system to generate informative features, which in turn are used by a machine learning classifier to identify the turbine availability states. The classifier is evaluated using K-fold cross-validation and robustness under increasing environmental noise levels. The main contributions of this work are the reduction in uncertainty and the association with real-time operating conditions, which enable optimal scheduling decisions. The baseline XGBoost classifier achieved an F1-score that increased after adding fuzzy-derived features. Comparative evaluation under noise-free and increasing noise levels demonstrates that the proposed framework consistently outperformed the baseline model while maintaining robust classification performance.