Inspection enhancement of wind turbine blades using novel signal-processing approaches

In this paper, a signal-processing approach is applied to enhance the inspection of wind turbine blades. This method starts with a time-frequency representation of signals, where a generalized time-bandwidth product method integrated with short-time Fourier transform is developed and applied. It is followed by the utilization of convex set to help justify the contour map such that the blade-condition monitoring can be better realized. Through test signals received from the sensors, it was found that the proposed approach would discriminate the damaged structure from the healthy one under several scenarios, thereby ensuring a higher efficiency of wind power generation.