Turbine loss model construction method based on deep learning and loss weight analysis

The invention discloses a turbine blade loss model construction method based on deep learning and loss weight analysis, and the method comprises the steps: firstly, carrying out the splitting of each loss in the blade loss, comparing with the prediction size of an existing model, and carrying out the correction of an existing turbine blade loss model; coefficients and items needing to be corrected and correction items needing to be added are obtained through analysis, and then a loss prediction model with a correction form is formed. And through comparison of turbine blade profile losses with different blade profile parameters, a blade profile parameter variable needing to be considered is found. And establishing a function relationship between a blade profile parameter variable needing to be considered and a coefficient (or item and a correction item needing to be added) needing to be corrected by utilizing an artificial neural network model, and substituting the function relationship into a loss prediction