Optimization of Stable Renewable Energy Generation through the Hybrid Use of Wind and Hydrogen Resources

Artificial intelligence–based methods, including big data analytics and machine learning, can be used to detect anomalies in the operating behaviour of wind turbines at an early stage. These methods can complement conventional remote monitoring and supervisory systems by identifying anomalies that might otherwise remain undetected. In doing so, they provide additional support for predictive maintenance and contribute to the reliable operation of wind farms. In summary, the early and reliable detection of changes in turbine operating behaviour is essential for preventing potential damage and limiting losses that may increase over time.

Neural-network models developed using input variables derived from historical data can be used to characterise the normal operating behaviour of a wind turbine. The algorithms can autonomously identify previously unknown dependencies and correlations within the data and construct the corresponding models, enabling autonomous modelling through machine learning. Measured values and operational data can therefore be continuously monitored and processed, allowing changes in turbine behaviour to be detected and anticipated. This approach can help mitigate the effects of the inherent variability of wind and contribute to more stable and reliable power generation.