A Machine-Vision-Based Hub Location Detection Technique for Installing Wind Turbines

Abstract

AbstractA position detection approach based on machine vision is developed to address the issue that the traditional wind turbine hub lifting procedure relies on the location of engineers and cannot provide quick feedback on the position information. A camera and a range sensor are mounted on the suspended object to enable real-time detection of the fan hub’s posture state in relation to the engine room. A monocular camera positioning technique based on circular features is created for this detection method. The relative pose coordinates of the hub are computed using the position data given by the range sensor. A filter for the range sensor is created based on the hoisting features of the actual working circumstances. The filter applies a filter to the data from the ranging sensor, considerably reducing measurement error and enhancing detection technique stability.