Cable-stayed bridge typical damage identification method

The invention discloses a cable-stayed bridge typical damage identification method. The method comprises the steps of S1, acquiring a data sample: establishing a cable-stayed bridge finite element numerical model under a typical damage working condition, and acquiring a data sample under a bridge intact and typical damage condition; S2, obtaining sensitive parameter indexes: analyzing the obtaineddata sample to obtain parameter indexes sensitive to typical damage; S3, training and testing a BP neural network: training and testing the BP neural network by utilizing parameter indexes sensitiveto typical damage, and evaluating and correcting the constructed BP neural network; and S4, damage identification: inputting the bridge actual monitoring data into the corrected BP neural network to realize bridge damage identification. The method has the advantages of being more reasonable in prediction mode, higher in application and popularization performance, high in damage recognition accuracy and the like. 本发明公开了一种斜拉