The bucket foundation is considered as a promising alternative to conventional foundations for offshore wind turbines (OWTs) due to its cost-effectiveness and high-reliability. The behavior of bucket foundation under static loading has been extensively investigated, while its long-term performance induced by the cyclic loading under the surrounding soil influence lack of study. This study aims to investigate the influence of one-way cyclic horizontal loading on the long-term performance of the wide-shallow bucket foundation (WSBF) model for offshore wind turbine in saturated sand by using single-gravity (1-g) model tests. Specifically, a series of cyclic experiments with varied loading conditions including load magnitude (P), load frequency (f), excitation height (H) and cycle number (N) were conducted to reveal the influence of loading conditions on dynamic characteristics and accumulated rotation of the WSBF model. Furthermore, the regression analysis method based on artificial neural network (ANN) model is proposed to determine the relationship between loading conditions and the long-term performance of the WSBF model. It is shown that the natural frequency rises and damping ratio decreases with the increase of cyclic loading number during the early cyclic stage. Afterwards, there is a declining or stable trend for the natural frequency and a rising trend for the damping ratio of the WSBF model. As for the accumulated rotation, more than 80% foundation rotation occurs in the first hundred cycles, whereas accumulated rotation for rest cycles tends to be small. Finally, multiple regression analysis and sensitivity analysis based on the ANN model are demonstrated to evaluate and predict the changes in long-term performance of the WSBF model with high accuracy in this study.

桶形基础由于其成本效益和高可靠性而被认为是海上风力涡轮机（OWT）常规基础的有前途的替代品。桶形基础在静态载荷下的行为已被广泛研究，而在周围土壤影响下循环载荷引起的长期性能缺乏研究。本研究旨在通过单重力（1-g）模型试验研究单向循环水平荷载对饱和砂中海上风力发电机的宽浅桶基础（WSBF）模型的长期性能的影响。具体来说，是一系列循环实验，其载荷条件不同，包括载荷大小（P），载荷频率（f），激发高度（H））和循环数（N）来揭示载荷条件对WSBF模型的动力特性和累积旋转的影响。此外，提出了一种基于人工神经网络模型的回归分析方法，确定了载荷条件与WSBF模型的长期性能之间的关系。结果表明，在早期循环阶段，固有频率随循环载荷数的增加而增大，阻尼比减小。此后，WSBF模型的固有频率有下降或稳定的趋势，而阻尼比则有上升的趋势。至于累积旋转，在前一百个循环中发生了超过80％的基础旋转，而其余循环的累积旋转趋于变小。最后，