Abstract:

With the rapid development of wind power industry, there are frequent accidents of wind turbines since design defects, manufacturing problems and inadequate maintenance. Collapse of units is one of the typical accidents, which can be avoided by measuring the inclination of the tower. In order to improve the low detection accuracy caused by the detection outliers and noise that cannot be eliminated by the least squares method, it proposes a random sample consensus algorithm which has a better immunity to noises and outliers. According to the testing results from a wind farm in Guazhou, Gansu, it is proved that the algorithm can be applied well in tower barrel inclination detection.

Key words:

wind turbine unit；tower barrel；inclination, random sample consensus algorithm, least square method, accuracy