This paper proposes a topology optimization and intelligent simulation modeling method for impellers aimed at high-efficiency hydrodynamics, addressing issues such as long design cycles, single optimization objectives, and the difficulty of adapting to complex operating conditions in traditional impeller design methods. By integrating an improved density method (SIMP) with the level set method, a multi-physics coupling topology optimization model is constructed, considering multi-objective constraints such as efficiency, stability, and cavitation performance. An improved MMA (Method of Moving Asymptotes) algorithm is employed to achieve the collaborative optimization of design variables. Using deep neural network to construct CFD proxy model significantly reduces the optimization cost and improves the design efficiency. The experimental verification shows that the optimized impeller is superior to the traditional design in key indexes such as hydraulic efficiency, lift and maximum stress, and the optimization time is greatly reduced. In addition, the optimized impeller shows good performance under multiple working conditions, and the prediction accuracy of the proxy model has also been verified. This method provides a new technical approach for efficient design of impeller and has important engineering application value.
Song et al. (Sun,) studied this question.