AI-based Power Production Models for Increased Wind Farm Efficiency

Wind energy is a clean and promising source of power, but making the most of it is not always easy. In a wind farm, multiple turbines work together, and their performance depends on how they interact with one another and with the surrounding environment. Factors such as hills, forest canopies, wind patterns between turbines (known as wakes), and even ice buildup on the blades can all affect how much energy is produced. 

 

Today, predicting a wind farm’s power output often requires long and complex computer simulations. In this project, we are exploring a faster, smarter approach using artificial intelligence (AI). By training AI models on real-world data, we aim to predict wind farm output more quickly and accurately, without relying on heavy computing power. 

 

We will use a special type of AI called Graph Neural Networks (GNNs), which are great at understanding systems where many parts are interconnected, like turbines in a wind farm. This approach could improve how wind farms are designed and operated, making wind energy more efficient, reliable, and affordable as we move toward a greener future.