On the Limitations of Machine Learning (ML) Methodologies in Predicting the Wake Characteristics of Wind Turbines

Abstract

Machine Learning (ML) algorithms have been more prevalent in recent years, and they are being used to tackle complicated issues across a broad range of fields. Wind energy is not an exception, as ML has recently been applied to wind turbine blade design, wake velocity and wake turbulence intensity prediction, and even wind farm optimization. The immense learning ability of ML models enables them to be trained to predict and regress a complex relationship with a high degree of accuracy. However, data for testing ML models often originate from the same rotor simulation used for training, with just slight variations in operating conditions. This research aims to investigate the generalizability of ML-based wake prediction models, i.e., whether ML can correctly predict wake properties using data from a different wind turbine that was not taken for training. This investigation's observation shows that a generalized ML wake model requires training data from multiple turbines with a wide range of operating conditions. In addition, advanced regularization, complex loss functions, and ML methods that focus on capturing the physics (such as Physics Informed Artificial Neural Networks (PINN) and symbolic regression) can be utilized.

Machine Learning-Based Approach to Wind Turbine Wake Prediction under Yawed Conditions

“…Moreover, machine learning algorithms have been integrated with physics-based models to yield hybrid methodologies that strive for increased generalization. The focus on the generalizability of these models has been a noteworthy avenue of investigation, aiming to ensure that machine learning-based wake models can predict properties across multiple turbines and varying operating conditions [65].…”

“…Nevertheless, while machine learning brings forth numerous advantages, it is crucial to exercise prudence. The limitations in machine learning approaches, particularly in the context of training data and the need for advanced regularization techniques, still present challenges that warrant further research [65].…”

Decision-Support Algorithm for Agronomic Practices: A Software Approach for Modeling Mechanized Tillage Planning

“…When developing the algorithm for decision support in tillage planning, a deliberate choice was made to utilize a compatibility matrices model instead of machine learning techniques. While machine learning is a powerful tool commonly employed for pattern recognition and prediction, it may not be the optimal approach for addressing the intricacies and interrelationships among multiple variables [2][3][4], typical of the work with natural resources. The complexity of these relationships necessitated a tailored solution that could effectively analyze and process the dependencies between soil types, crop types, machinery options, and other relevant factors.…”