Tower of wind turbine generator is used as bracing structure of generator set

and its perpendicularity has strong sensitivity to its safety and stability. Affected by long-term natural and human factors

the tower suffurs surface damage and structural inhomogeneity. As a result

the conventional monitoring methods will generate abnormal errors

which directly leads to distortion of the measurement results. To solve this problem

the authors use the robust estimation theory to carry out perpendicularity analysis on wind power towers for the first time. Moreover

combined with least square method

the center and radius of the measured value of the section of the tower tube are fitted

and by using Huber selecting weight iteration method

it will gradually eliminate the influence of abnormal measuring points such as gross errors and accidental errors in the measured value series. Finally

this method is verified by an instance. The results show that

the robust least-square circle fitting has ability to resist errors

which can effectively avoid the influence of abnormal measurement values such as gross errors and accidental errors

and improve the reliability of fitting parameters and perpendicularity analysis. This method has the advantages of clear logic

simple programming method and practical value. The research results provide a reference for the perpendicularity analysis of the tower of wind turbine generator and similar towers.