The hydrodynamic performance inside the recirculating aquaculture tank is critical for the welfare of aquaculture. It is undoubtedly a preferable alternative to achieve a conducive flow environment inside the aquaculture tank through optimizing the geometry structure of the aquaculture tank. In this study, combining the computational fluid dynamics (CFD) approach and multi-objective genetic algorithm (MOGA), these two parameters, namely α (R/L) and β (C/L), are assessed and optimized based on the average flow velocity and flow velocity uniformity inside the square arc angle aquaculture tank. Physical experiments are set up to verify the accuracy of the numerical models and predicted results. The results indicate that the variation of both α and β affects the flow velocity and vortex distribution. However, the sensitivity of flow velocity uniformity and average flow velocity to α (74.49% and 91.14%) is much higher than that of β (−24.12% and −23.68%). It demonstrates that increasing α can significantly improve the average flow velocity and flow velocity uniformity inside the aquaculture tank compared to β. In addition, according to MOGA optimization results, taking the maximum flow velocity uniformity and average flow velocity as the optimization goals, the final optimization scheme is α = 0.396 and β = 0.074. Finally, combining CFD and MOGA methods can optimize the aquaculture tank structure accurately and efficiently in a comprehensive and systematic approach with relatively minimal computational cost.

循环水养殖池内的水动力性能对水产养殖的福利至关重要。通过优化养殖池的几何结构来实现养殖池内部的有利流动环境无疑是一种更可取的选择。在本研究中，结合计算流体动力学（CFD）方法和多目标遗传算法（MOGA），这两个参数，即α（R / L）和β（C / L），根据方弧角养殖池内的平均流速和流速均匀性进行评估和优化。设置物理实验验证数值模型和预测结果的准确性。结果表明， α和β的变化都会影响流速和涡流分布。然而，流速均匀性和平均流速对α（74.49%和91.14%）的敏感性远高于β（-24.12%和-23.68%）。说明增加α与β相比可以显着提高养殖池内的平均流速和流速均匀性. 另外，根据MOGA优化结果，以最大流速均匀性和平均流速为优化目标，最终优化方案为α  =0.396，β  =0.074。最后，结合 CFD 和 MOGA 方法可以以相对最小的计算成本以全面和系统的方法准确有效地优化水产养殖池结构。