Abstract: Owing to the large volume and complex structure of high earth-rockfill dams, identifying their modal parameters is challenging, posing difficulties to subsequent dynamic analysis. Based on earthquake measurement data, this paper presents a fuzzy C-means clustering algorithm that is based on modal stability functions, and constructs a new method of automatic modal parameter identification, namely MSFCM-SSI, by combining the algorithm with a covariance-driven stochastic subspace identification method. This new method achieves accurate and automatic identification of dam modal parameters, avoiding errors associated with traditional manual selection. Its effectiveness and accuracy are validated through application to a two-degree-of-freedom numerical model and an earth dam￠s two-dimensional finite element model. It is then used to identifying the modal parameters of the Lianghekou high-core rockfill dam, demonstrating its applicability and advantage in practical seismic analysis of high earth-rockfill dams.

Key words: high earth-rockfill dam, modal parameter, stochastic subspace, stability diagram, fuzzy C-means clustering