Commercial Aircraft Corporation of China (China)

Low-level wind shear is a frequent hazardous phenomenon at airports in the Xinjiang region of China, mainly due to complex terrain and highly variable weather conditions. It poses a significant risk to aircraft operations, particularly during take-off and landing. In this study, Doppler wind lidar observations are used to detect and identify low-level wind shear in the vicinity of airports, with a focus on improving the performance of existing identification algorithms under complex terrain conditions. Several commonly used wind shear detection algorithms are implemented, evaluated, and further refined. Based on their complementary strengths, a joint warning algorithm is developed to provide more reliable wind shear alerts. In addition, machine learning methods are explored to directly extract wind shear signals from raw lidar data, aiming to achieve faster detection without relying on full wind field retrieval. The results show that the joint warning algorithm clearly improves warning performance compared to individual algorithms, with fewer false alarms and missed events. The machine learning approach also demonstrates promising capability for rapid wind shear identification. These results suggest that combining multi-algorithm warning strategies with data-driven methods can effectively enhance future airport wind shear warning systems in the Xinjiang region. Further improvements are expected by training machine learning models with expanded libraries of representative wind shear cases.

Artificial intelligence (AI) models have demonstrated advancements in computational efficiency and forecast accuracy relative to the Numerical Weather Prediction (NWP), but they are unable to fully represent high-dimensional atmospheric dynamics. Thus, some AI-NWP coupled frameworks have been proposed, such as integrating AI-driven boundary conditions with numerical models to leverage the strengths of both approaches. However, in this coupled framework, ensemble forecasts and associated error propagation and energy dynamics remain under-explored. In this study, an AI-NWP coupled system that also uses the stochastic kinetic energy backscatter scheme (SKEBS) to generate ensemble forecasts is established. Ensemble simulations of Typhoon Yutu (2018) are carried out with the Weather Research and Forecasting (WRF) model employing Pangu-Weather and FuXi forecast data as boundary forcing. The results show that the ensemble WRF_Pangu (WRF_FuXi) improved Yutu’s track forecast by 67% (50%) compared to the traditional physics-based WRF_GFS (Global Forecast System), and reduced its intensity underestimation by about 67% relative to their AI global counterparts. Nonetheless, WRF_FuXi and WRF_Pangu exhibited limited ensemble spread and linear error growth, reflecting deterministic tendencies. Comparison of global and regional experiments show that Pangu-Weather is more physically constrained and thus better aligned with the WRF model for regional applications, while the adaptation of FuXi to the regional model is less robust. Spectral analysis revealed that AI-derived boundaries introduced excessive small-scale energy and underestimated larger-scale energy. The regional model WRF acted as a “conveyor belt”, propagating additive small-scale energy upscale, ultimately overwhelming the stochastic perturbations for ensemble generation. These findings underscore the need to incorporate more physical features into the AI-derived boundary conditions for ensemble forecasting.

Civil aviation VHF communication is safety-critical, yet operational links are routinely disturbed by atmospheric effects, aging hardware, and electromagnetic interference. The resulting anomalies are typically weak, intermittent, and extremely rare, which makes real-time detection difficult under strong temporal dependence and severe class imbalance. We propose an end-to-end framework that couples (i) a learnable kernel projection for adaptive nonlinear feature extraction, (ii) a differentiable relevance–redundancy objective for feature refinement, and (iii) conditional temporal generation to augment minority anomaly patterns. A lightweight CNN–LSTM head is used for streaming inference. Training uses a mixture of operational anomalies and simulated degradation scenarios, while evaluation is conducted using operational data only. Experiments on 1.2 million VHF frames collected from real flight operations and ground station monitoring achieve an F1-score of 0.947, ROC-AUC of 0.972, and PR-AUC of 0.968, with an average inference latency of 34.7 ms.

With the rapid development of the low-altitude economy, the intelligence level of aircraft urgently needs improvement. In emergency landing scenarios, traditional guidance systems are prone to communication failures, adverse weather, or complex environmental interference, posing significant safety risks. To address the high-precision pose estimation requirements for autonomous aircraft landing, this paper proposes a monocular vision-based sparse keypoint parameterization method. Our approach employs a deep neural network to rapidly detect landing-area keypoints, establishes 2D-3D sparse correspondences, and leverages Perspective-n-Point (PnP) for real-time pose solving, overcoming the limitations of traditional systems that rely on prior calibration or cooperative markers. Experiments demonstrate that our method achieves robust pose estimation even under extreme conditions (e.g., signal interference, low visibility), providing reliable technical support for emergency landings. A computationally efficient and interference-resistant sparse keypoint representation for targets. And a cooperative-marker-free monocular vision landing guidance system, significantly enhancing landing safety in complex scenarios.