his dataset includes two parts. The first part is three sets of physically measured blade-root flapwise bending moments on three respective turbines, courtesy of Riso-DTU (Technical University of Denmark). The basic characteristics of the three turbines can be found in Table 10.1 of the Data Science for Wind Energy book. These datasets include three columns. The first column is the 10-min average wind speed, the second column is the standard deviation of wind speed within a 10-min block, and the third column is the maximum bending moment, in the unit of MN-m, recorded in a 10-min block. The second part of the dataset is the simulated load data used in Section 10.6.5 of the same book. This part has two sets. The first set is the training data that has 1,000 observations and is used to fit an extreme load model. The second set is the test data that consists of 100 subsets, each of which has 100,000 observations. In other words, the second dataset for testing has a total of 10,000,000 observations, which are used to verify the extreme load extrapolation made by a respective model. Both simulated datasets have two columns: the first is the 10-min average wind speed and the second is the maximum bending moment in the corresponding 10-min block. While all other datasets are saved in CSV file format, this simulated test dataset is saved in a text file format, due to its large size. The data simulation procedure is explained in Section 10.6.5.

{"references": ["Ding, Y. (2019) Data Science for Wind Energy, Chapman & Hall/CRC Press, Boca Raton, FL"]}