Stereo vision-based health monitoring method for wind turbine towers

Abstract

Wind turbines are subjected to complex aerodynamic loads, which results in potential sway and tilt, even causing the collision between blades and the tower. Therefore, it is imperative to monitor the condition of the tower to ensure safety. In this paper, we propose a method to monitor the health of wind turbine towers based on stereo vision. Our approach extracts the tower central axis from the image based on the structural information. Then the tower’s sway magnitude is measured by binocular intersection, and the tilt angle is computed by exploiting the angle information provided by an inclinometer. Moreover, we detect blade tips using their geometric shape, which can be used to estimate the clearance distance between blades and the tower. The proposed method can simultaneously monitor the tower tilt and sway, while also identifying instances of blade collision with the tower. Experiments demonstrate the accuracy and feasibility of the designed system.