The invention provides an anomaly identification method and device of a wind power generator set. The identification method comprises the following steps: obtaining a conventional wind speed and powerdata of normal operation of the wind power generator set; establishing a wind speed power coordinate system according to the conventional wind speed and the power data; carrying out grading processing on the wind speed power coordinate system by a pre-set interval, so as to obtain wind speed power data of a plurality of grades; carrying out fitting on the wind speed power data of each grade in the wind speed power data of the plurality of grades to obtain a wind speed power model; according to the wind speed power model, calculating to obtain an average value of power absolute errors of the wind speed power data of each grade in the wind speed power data of the plurality of grades and a standard difference of the power absolute errors; determining an alarming threshold value according tothe average value of the power absolute errors and the standard difference of the power absolute errors; substituting the power data of the wind power generator set to be detected into the wind speedpower model and calculating the power absolute error of the wind power generator set to be detected; comparing the power absolute errors of the wind power generator set with the alarming threshold value and outputting state information of the wind power generator set to be detected according to a comparison result.