Normalized vibration signals of blades with different cracked zones in a wind turbine in operation

The dataset includes normalized vibration signals acquired directly at the root of each of three wind turbine blades of a low-power rotating wind turbine. The tests were conducted at the same wind speed. The wind turbine blades have different health conditions; the cases considered are healthy blades and cracked blades in one of three cross-sectional areas: root, middle, or tip. There are 10 cases, with 7680 files each. The dataset was collected at a sampling rate of 1 kHz, with 500 samples per channel. The normalization was performed using the absolute maximum (Max Absolute) Scaling Method. The technique was implemented in Python using the scikit-learn (version 1.4) library.