The tripod bucket jacket foundation is proven to be a practicable solution for offshore wind turbines (OWTs) to withstand huge environmental loads in deep water. This paper presents model tests for a scaled tripod bucket jacket foundation with reference to a prototype applied in China to obtain its lateral load bearing behavior in medium-dense sands. Extended finite element analyses were conducted by ABAQUS to compare anti-overturning responses for the tripod bucket foundation in both sand and soft clay, and the influences of loading direction and aspect ratio were also taken into account. The results showed that the failure modes of the laterally loaded tripod bucket foundation are the pull-out of the windward bucket in sand and the settlement of the leeward bucket in soft clay, respectively. Thus, the unfavorable loading direction of the foundation changes with soil type. It is also shown that the bearing capacity for the foundation in soft clay will be enhanced more effectively as the bucket diameter increases. Instead of the rotational soil resistance resulting from the rotation of the bucket, the vertical soil resistance governs the anti-overturning bearing capacity of a tripod bucket foundation. As the tilt created by the overturning moment rises, the rotational stiffness of the foundation dramatically declines.

This study investigates class imbalance in motor fault diagnosis. Fault samples, especially those at different severity levels, are often much fewer than healthy samples. To address this issue, a self-attention guided Wasserstein conditional GAN with gradient normalization (SWGAN) is proposed. The method is based on synchronized three-phase current and vibration measurements. It separately generates label-conditioned current spectra and vibration spectra to supplement minority fault classes. Self-attention is used to capture long-range spectral dependencies. Gradient normalization is introduced to improve adversarial training stability. The generated current and vibration spectra are then fused at the feature level and fed into a stacked autoencoder (SAE)-based multi-modal classifier. Experiments were conducted on a PMSM stator fault dataset and a variable-speed three-phase asynchronous motor dataset. On the PMSM dataset, SWGAN achieved highest accuracies of 98.90% and 97.81% under two fault-category imbalance settings. On the variable-speed motor dataset, the proposed method achieved accuracies of 98.10% and 97.65%, respectively. These results show that SWGAN can provide effective supplementary samples for minority fault classes. They also indicate that the proposed method improves diagnostic performance under both fixed-speed and variable-speed conditions.