A scanning lidar measures a projection of the actual wind speed to the line-of-sight. Reconstruction of the original field is possible via various retrieval algorithms, especially if several lidars are deployed in the area. However, this may not be the case due to financial constraints. Additionally, conventional algorithms like Volume Velocity Processing (VVP) do not perform well for non-homogeneous flow fields, e.g., smooth wind turbine wakes in a retrieved field. A 2D-VAR algorithm using VVP as an intermediate step was suggested allowing the retrieval for consecutive scans at a low elevation angle. We validate this algorithm for scans containing prominent wakes under elevation angles within 5° using FINO1 mast and supervisory control and data acquisition (SCADA) data. We show that prominent wakes scanned at the hub height introduce substantial disturbances in the VVP retrieval output and affect the retrieved field. The unwanted disturbance may be overcome by masking the wakes. In point measurements, the wind speed magnitude is retrieved with an acceptable accuracy, but the wind direction appears to be sensitive to the weights and initial guess chosen. According to our sensitivity analysis, the weight assigned to the radial velocity residuals affects the outcome most, while other weights act as corrections.

Bolt looseness may occur on wind turbine (WT）blades exposed to operational and environmental variability conditions, which sometimes can cause catastrophic consequences. Therefore, it is necessary to monitor the loosening state of WT blade root bolts. In order to solve this problem, this paper proposes a method to monitor the looseness of blade root bolts using the sensors installed on the WT blade. An experimental platform was first built by installing various blade sensors for monitoring bolt looseness. Through the physical experiment of blade root bolts looseness, the response data of blade sensors were then obtained under different bolt looseness degrees. Afterwards, by analyzing the sensitivity of the response signal to the looseness of blade root bolts, the types and number of sensors that can judge the looseness of blade root bolts were determined. Finally, the multi-domain sensitive features of response signals were fused to construct a hybrid domain feature set of bolt looseness. The LightGBM classification algorithm was applied to identify different bolt loosening states for this hybrid domain feature data. The identification results of experimental data showed that the proposed method can accurately determine the loosening state of single or multiple WT blade root bolts.

Vortex generators (VGs) have been widely applied to wind turbines thanks to their potential to increase aerodynamic performance. Due to the complex inflow perceived by a rotor and the proneness to flow separation, VGs on wind turbines usually experience highly unsteady flow. While there are models that exist to simulate the steady effects of VGs, we lack a fast and efficient tool to model the unsteady performance of airfoils equipped with VGs. This paper adopts an unsteady double-wake panel model with viscous-inviscid interaction developed to simulate a vertical axis turbine in dynamic stall, adding the capability of predicting the dynamic aerodynamic performance of VG-equipped airfoils. The results of a series of steady and unsteady cases of an airfoil with different VG configurations in various pitch motions in free and forced transition are verified against experimental data. Results show that the double wake model offers sufficient accuracy results compared to experimental data to claim the model’s validity in a preliminary evaluation of an airfoil’s capability to prevent stall with VGs. While a few limitations are still identified for improvement, the model’s accuracy in predicting the transition location, separation and reattachment, and drag forces into future development.