In electronic warfare, jammers are capable of transmitting interference types that have not been previously encountered during radar agent training, thereby rendering radar anti-jamming models ineffective. Due to the short duration of adversarial engagements, it is challenging to acquire sufficient data on unknown jamming conditions to meet the imitation learning requirements of radar anti-jamming models. To address this issue, we propose an imitation learning method for radar anti-jamming based on signal feature retrieval, specifically designed for few-shot jamming data. The proposed method first constructs a signal feature embedding model using sufficient offline reinforcement learning interaction data. Under few-shot unknown jamming scenarios, the proposed method directly encodes the received jamming I/Q observations and retrieves high-value state–action experiences from historical radar–jammer interactions, without explicitly identifying the jamming category. The retrieved experiences are subsequently incorporated into the policy optimization process to generate radar anti-jamming actions, while newly collected interaction data are used for further task-specific adaptation through reward feedback. Experimental results demonstrate that the proposed method can generate effective low-dimensional representations for large-scale offline state–action pairs, achieving a waveform reconstruction similarity of 96.18%. The feature retrieval model constructed on this basis, which is adapted to few-shot unknown jamming scenarios, improves the average detection probability by 57.05% with 150 samples. In addition, the proposed method outperforms conventional approaches in terms of both environmental robustness and resistance to knowledge forgetting.
Energy is the binding constraint on unmanned aerial vehicle survey missions, yet most coverage path planners commit to a single sweep pattern regardless of region shape. Our central contribution is to select decomposition granularity and the per-cell coverage pattern by a single physics-based energy criterion. A boustrophedon cellular decomposition with a controlled-concavity merge sets granularity; for each sub-polygon, the planner evaluates a minimum-width lawnmower and a shape-mimicking spiral and keeps the cheaper, with the undecomposed single-sweep solution always in the search space. On 50 benchmark polygons, at a 90% coverage floor, the hybrid reduces energy over a spiral baseline by a median 4.8% (ratio of means 11%) and is never dearer than either fixed pattern at floors of 90–99%. Which pattern is cheaper depends on geometry—lawnmower and spiral are not separable in aggregate (Wilcoxon p=0.47). A boundary-reflective zigzag for double coverage achieves 46% fewer turns and 18% less energy than boustrophedon 2×. The hybrid keeps the lowest mean energy under ±20% parameter perturbation and on larger fields with wider row spacing.
Thin-walled carbon fiber-reinforced polymer (CFRP) members with closed lenticular cross-sections are suitable for lightweight load-bearing and deployable aerospace structures. Their axial buckling may be governed by competition between finite-length local or cross-sectional deformation branches and long-member global flexural branches. This competition increases the computational cost of parametric sweeps and complicates rapid identification of the transition length. This study develops a combined Rayleigh–Ritz–Euler screening framework to address this challenge. A finite-length Rayleigh–Ritz (RR) model predicts local and mixed buckling branches and the governing half-wave number, while an Euler model provides the global flexural buckling load scale. The intersection of the predicted load–length curves defines an engineering transition length, and shell finite element (FE) analyses are used to check selected configurations. The RR model incorporates an effective closure coefficient for the geometric stiffness, calibrated solely on the baseline configuration (BASE). Against Abaqus linear buckling results, excluding BASE, RR and Euler yield root-mean-square relative errors of 11.6% and 1.3% in the local and global buckling regimes, respectively. Across all 13 configurations, including BASE, the engineering transition length has a mean absolute relative error of 2.4% and a maximum of 11.1%. The framework captures transition-length trends under the investigated variations in cross-sectional scale, thickness, and layup, but shows larger deviations for variations in the circular-arc subtended angle. Within the tested parameter range, it supports preliminary screening of candidate designs and identification of parameter regions for targeted FE verification. However, it cannot replace high-fidelity FE analysis or experiments as the basis for final design decisions.
Rapid prediction of cavity sound pressure spectra is essential for the acoustic design of high-speed vehicles. A spectral database is built from 20 flow conditions (4 Mach numbers × 5 angles of attack) and eight cavity measurement points, giving 160 condition–point combinations, each densely sampled at 1147 amplitude–frequency points over 6–7000 Hz. Four models—RF, XGBoost, LightGBM and SVM—are compared on two prediction tasks, flow-condition and spatial-position prediction, before and after Bayesian optimization. Optimization markedly improves in-distribution accuracy and reduces dominant-frequency amplitude errors under unseen conditions, but reverses the ranking: RF and LightGBM dominate at baseline, whereas XGBoost ranks first after tuning, attaining 0.76 dB in the flow-condition task and staying on a par with RF (0.09 dB apart) in the spatial-position task. Under unseen conditions the SVM degenerates to a constant output of 117.11 dB, essentially the training-target mean, owing to the locality of the RBF kernel—a structural limitation that tuning cannot remedy. The direction of the peak-amplitude bias is not uniform: systematic underestimation occurs in the spatial-position task, whereas after tuning XGBoost overestimates the peak at Ma = 1.5, the condition with the highest measured amplitude. No single model is optimal across all scenarios.
Reliable visual delineation of ice accretion on aircraft wing leading edges remains challenging because accreted ice often exhibits low visual contrast, translucent and irregular morphology, and ambiguous boundaries. To address these challenges, this study proposes ICE-UNet, a lightweight semantic segmentation network for pixel-wise leading-edge icing detection. The network adopts a compact encoder–decoder architecture and integrates three functional modules: Residual Depthwise Separable Convolution with Batch Normalization and ReLU Activation (Res-DSCBR), Depthwise-Separable Receptive Field Blocks (DSCRFBs), and Convolutional Block Attention Modules (CBAMs). This design reduces model complexity while enhancing multi-scale contextual representation and attention-guided feature refinement. For empirical validation, a dataset of 5000 annotated image–mask pairs was constructed from 1300 original images comprising wind-tunnel video frames and post-icing photographs of the NACA 0020 and Boeing 117 airfoil models. Following augmentation, the 5000 pairs were randomly divided at the image level into training, validation, and test subsets in a 3:1:1 ratio. Experimental results show that ICE-UNet achieves 96.60% accuracy, 93.07% intersection over union (IoU), 96.36% precision, and 96.44% recall, with only 0.76 M parameters and 16.19 GFLOPs. Compared with the baseline U-Net, ICE-UNet improves IoU by 10.33 percentage points while reducing the parameter count and computational cost by 95.6% and 90.0%, respectively. Photometric perturbation tests further indicate that the model maintains stable segmentation performance under moderate brightness and contrast variations, although severe contrast degradation remains a limiting condition. TensorRT-based FP16 acceleration reduced the inference latency from 55.21 ms to 13.69 ms per image on a Tesla V100 GPU without measurable accuracy degradation. Additionally, a MobileNetV3 pre-screening module was integrated to effectively suppress false-positive activations on clean surfaces during continuous monitoring. Collectively, these findings show that ICE-UNet achieved the highest IoU among the ten segmentation models evaluated on this wind-tunnel dataset while retaining a compact architecture.
