Abstract: To address the challenge of accurately establishing a dynamic model for thin-walled workpiece systems in mirror milling, where complex boundary conditions from fixtures and active supports introduce significant inaccuracies, a high-fidelity dynamic modeling method based on S4R shell elements and a genetic algorithm is proposed to predict the dynamic characteristics of thin-walled workpiece-support systems in mirror milling under various clamping conditions. Based on a comparative mesh convergence analysis between S4R shell and C3D8R solid elements, the S4R shell element is selected to construct the initial dynamic model of the system, and the fixtures and the support equipment are modeled as equivalent spring elements to impose the boundary conditions. Using the error between the experimentally measured and simulated frequency response functions (FRFs) of the workpiece-support system as the objective function, a genetic algorithm is employed to identify the optimal values for parameters such as equivalent spring stiffness and modal damping ratios. The results demonstrate that the FRFs computed from the optimized shell element finite element model exhibit excellent agreement with the experimental results, enabling a more accurate prediction of the system's FRFs. The proposed method provides the foundational initial model for subsequent predictions of time-varying dynamic parameters, serving as a prerequisite and methodological support for the high-precision stability prediction and offline process optimization of the thin-walled workpiece mirror milling process.