This method uses the existing wind turbine operation data, environmental data and meteorological data to realize the prediction and quantitative evaluation of wake flow through non-intrusive mode. In this method, the characteristic factors affecting the wake are extracted and the wake prediction classification tree is constructed. The feature classification tree is used to determine whether the wind turbine is affected by the wake. This method uses the classification tree of benchmarking wind turbine to determine the latest benchmarking wind turbine and obtain health data. The wind speed under the influence of the wake is calculated according to the Gaussian model, and the wind speed loss rate is calculated to get the loss factor. According to the loss factor, the operation health degree of wind turbine under the influence of wake is quantified, which provides the decision basis for the yaw strategy of wind turbine.