abbey2017 / wind-energy-analytics

Physics-guided data-driven solutions for the wind energy industry
BSD 3-Clause "New" or "Revised" License
22 stars 8 forks source link

Develop a module to estimate the expected power for each data point using binning method. #18

Open abbey2017 opened 3 years ago

abbey2017 commented 3 years ago

The returned models from this module can be used to estimate the expected power from a turbine or wind farm given the wind speed vs power data from normal operating conditions.

abbey2017 commented 3 years ago

The goal is to develop a module which returns the expected power for each data point using binned operational power curve based on normal operating data of wind turbines.

abbey2017 commented 2 years ago

The module will accept training and test data from users.

The training data will be used to create normal data points using the power curve filtering module. Next, the expected power from test and training (optional) data will be estimated using different methods.

The first method will create a benchmark by binning the normal data points from the training data. Other methods may use a machine learning approach and also leverage other predictors like wind direction.