As an essential subsystem of electrified railway operation and maintenance, intelligent detection of catenary support components still faces several critical challenges: (1) the number of abnormal (negative) samples for components is severely limited; (2) component anomalies are highly diverse and exhibit heterogeneous visual characteristics; and (3) existing models generally show unsatisfactory detection performance when confronted with previously unseen anomaly types. To address these issues, this paper proposes a novel few-shot anomaly detection model for catenary components, termed BCLIP-ADer, built upon a Bayesian prompt contrastive vision-language pretraining framework. Specifically, a Bayesian prompt flow module (PFM) is designed to regularize the text prompt space via the jointly learned image-specific feature distribution (ISFD) and image-agnostic feature distribution (IAFD), thereby mitigating the degradation in detection performance on unseen component anomalies. Monte Carlo sampling over these learned distributions is further employed to generate diverse text prompts, leading to more comprehensive coverage of the prompt space. In addition, a cross-modal feature refinement module (CFRM) is designed to more effectively align dynamic text embeddings with fine-grained image features, thus enhancing anomaly detection at the component level. Finally, extensive experiments conducted on a UAV-based catenary dataset (CSCUD) demonstrate the effectiveness and superiority of the proposed approach. Specifically, the proposed method achieves I-AUROC/I-AP/I-F1_max scores of 94.2/93.2/93.1 under few-shot conditions.

Potato vigour, the growth potential of seed potatoes, is a key agronomic trait that varies significantly across production fields due to factors such as genetic background and environmental conditions. Seed tuber microbiomes are thought to influence plant health and crop performance, yet the precise relationships between microbiome composition and potato vigour remain unclear. Here we conducted microbiome sequencing on seed tuber eyes and heel ends from 6 potato varieties grown in 240 fields. By using time-resolved drone imaging of three trial fields in the next season to track crop development, we were able to link microbiome composition with potato vigour. We used microbiome data at varying taxonomic resolutions to build random forest predictive models and found that amplicon sequence variants provided the highest predictive accuracy for potato vigour. The model revealed variety-specific relationships between the seed tuber microbiome and next season’s crop vigour in independent trial fields. With a coefficient of determination value of 0.69 for the best-performing variety, the model accurately predicted vigour in seed tubers from fields not previously included in the analysis. Moreover, the model identified key microbial indicators of vigour from which a Streptomyces, an Acinetobacter and a Cellvibrio amplicon sequence variant stood out as the most important contributors to the model’s accuracy. This study shows that seed potato vigour can be reliably predicted based on the microbiota associated with seed tuber eyes, potentially guiding future microbiome-informed breeding strategies.

Due to its long-span structure and large flexibility, an electrified railway catenary is very sensitive to environmental wind load, especially the time-varying stochastic wind, which may lead to a strong forced vibration of contact line and deteriorate the current collection quality of the pantograph-catenary system. In this paper, in order to study the windinduced vibration behavior of railway catenary, a nonlinear finite element procedure is implemented to construct the model of catenary, which can properly describe the large nonlinear deformation and the nonsmooth nonlinearity of dropper. The spatial stochastic wind field is developed considering the fluctuating winds in along-wind, vertical-wind, and cross-wind directions. Using the empirical spectra suggested by Kaimal, Panofsky, and Tieleman, the fluctuating wind velocities in three directions are generated considering the temporal and spatial correlations. Based on fluid-induced vibration theory, the model of fluctuating forces acting on catenary are developed considering the spatial characteristics of catenary. The time- and frequency-domain analyses are conducted to study the wind-induced vibration behavior with different angles of wind deflection, different angles of attack, as well as different geometries of catenary. The effect of spatial wind load on contact force of pantograph-catenary system is also investigated.