An improved Eulerian–Lagrangian agent method framework for simulating fish behavior in fishways

Fishways serve as critical migration corridors for fish, and their design effectiveness directly impacts ecological conservation and resource management. This study develops a fish trajectory simulation model based on the Eulerian–Lagrangian agent method framework to understand better fish movement behavior within fishways and support hydraulic optimization. The model comprehensively considers fish responses to flow velocity and direction, improves the fish perception range, and incorporates social behaviors (e.g., following). Model performance was validated by comparing simulated trajectories with experimental data, using metrics such as linear regression analysis, root mean square error, and mean absolute error (MAE). Results indicate that the model achieves an MAE range of 0.024–0.043 m, with the maximum MAE accounting for only 8% of the fishway width. The simulated trajectories show high consistency with experimental trajectories regarding correlation, and the errors are within acceptable limits, demonstrating that the model can reasonably predict fish trajectory distribution within fishways.