Jing Xin

Unmanned aerial vehicles (UAVs) equipped with onboard processors enable edge intelligence for smart-grid Internet of Things (IoT) inspection, yet insulator defect detection from aerial imagery remains challenging due to cluttered backgrounds, small targets, and strict latency constraints. Compared with handcrafted methods that rely on manual feature design and parameter tuning, this article presents an end-to-end improved Faster R-CNN detector whose feature extractor is redesigned via convolutional feature reuse (CFR) and a two-layer multiscale fusion scheme with a deeper extraction path, preserving fine-grained cues while limiting model complexity. Experiments on Chinese Power Line Insulator dataset (CPLID) and a newly constructed Glass Insulator dataset (GID) achieve mAP@0.5 values of 97.75% and 96.79%, respectively, which is 7.1% higher than a standard Faster R-CNN (VGG16) baseline on CPLID while reducing the model size to 58 MB. On a UAV system (DJI M300 + H20T + Jetson Xavier NX), TensorRT acceleration runs at 8.5 frames/s, outperforming Faster R-CNN (3.2 frames/s) and YOLOv11 (5.3 frames/s) under the same edge setting. Outdoor flight tests on 330-kV transmission lines (TLs) further validate field readiness, supporting low-latency and bandwidth-efficient UAV inspection for smart-grid IoT maintenance.

Insulators are essential components in high-voltage transmission lines and require regular inspection to ensure reliable power delivery. Traditional manual inspection methods are inefficient and labor intensive, highlighting the need for intelligent and automated solutions. In this study, we propose a missing insulator detection method that integrates Unmanned Aerial Vehicle (UAV) imaging with deep learning techniques. Firstly, an improved Faster Region-based Convolutional Neural Network (Faster R-CNN) is employed to detect and localize insulators in aerial images. Secondly, the localized insulators are segmented using an improved U-Net to reduce background interference. A bounding box regression approach is adopted to obtain the minimum enclosing rectangles, and the insulators are aligned vertically. Adaptive thresholding is then applied to extract binary images of the insulators. These binary images are further transformed into defect curves, from which missing insulators are identified based on curve distribution. To address the limited availability of labeled samples, a transfer learning-based strategy is adopted to improve model generalization. A dataset of glass insulators was collected using a DJI M300 UAV equipped with an H20T camera along a 330 kV overhead transmission line. On the collected UAV insulator dataset, the proposed method achieved an AP@0.5 of 99.85% and an average IoU of 88.56% for insulator string detection, while the improved U-Net achieved an mIoU of 89.73% for insulator string segmentation. Outdoor flight experiments further verified performance under varying backgrounds and illumination conditions in our UAV inspection scenarios.

With the widespread application of unmanned aerial vehicles (UAVs), intrusion detection based on wireless communication traffic has become increasingly critical. While deep learning (DL) models have significantly enhanced anomaly detection accuracy, recent research indicates that they remain vulnerable to adversarial sample attacks, which pose severe threats to the open communication networks used by UAV systems. To address this challenge, an enhanced UAV intrusion detection framework is proposed for defending against adversarial samples, integrating generating adversarial samples with adversarial networks (AdvGAN) and decision boundary reconstruction. Specifically, a convolutional neural network (CNN) was first developed as the baseline intrusion detection model to detect anomalous traffic in UAV wireless communication. Subsequently, to overcome the gradient-dependent perturbation limitations of traditional adversarial attack methods, AdvGAN is employed to analyze real monitored traffic data and generate semantically consistent adversarial samples that retain the intrinsic characteristics of authentic network flow. Finally, a hybrid adversarial-real data training strategy is proposed to dynamically reconstruct the model’s intrusion decision boundaries. This strategy augments the training dataset with AdvGAN-generated adversarial samples, enabling the model to adaptively adjust decision thresholds through iterative backpropagation. On a real-world dataset, the superior performance of the proposed method is verified: it maintains high detection accuracy for legitimate traffic while substantially reducing the success rate of adversarial attacks. Experimental results confirm the stealthiness of generated adversarial samples and the robustness of the enhanced detection model, validating the effectiveness of the proposed framework.

This paper focuses on the linear quadratic control (LQC) design of systems corrupted by both stochastic noise and bounded noise simultaneously. When only of these noises are considered, the LQC strategy leads to stochastic or robust controllers, respectively. However, there is no LQC strategy that can simultaneously handle stochastic and bounded noises efficiently. This limits the scope where existing LQC strategies can be applied. In this work, we look into the LQC problem for discrete-time systems that have both stochastic and bounded noises in its dynamics. We develop a state estimation for such systems by efficiently combining a Kalman filter and an ellipsoid set-membership filter. The developed state estimation can recover the estimation optimality when the system is subject to both kinds of noise, the stochastic and the bounded. Upon the estimated state, we derive a robust state-feedback optimal control law for the LQC problem. The control law derivation takes into account both stochastic and bounded-state estimation errors, so as to avoid over-conservativeness while sustaining stability in the control. In this way, the developed LQC strategy extends the range of scenarios where LQC can be applied, especially those of real-world control systems with diverse sensing which are subject to different kinds of noise. We present numerical simulations, and the results demonstrate the enhanced control performance with the proposed strategy.

This article proposes a distributed intelligent formation control method for mobile robots via multisource perception. Multirobot formation control is challenging in unknown environments due to the lack of global pose information and the susceptibility of visual tags to disturbances. To cope with the constraints imposed by data, a novel deep reinforcement learning (DRL) flexible formation motion control framework is constructed. The followers compute motion decisions based on local multisource perception data, autonomously achieving shape formation and formation maintenance. Each robot optimizes the formation by predicting the pose of its leader and reduces the average path length of the formation. This framework designs a multistage DRL incremental training method for pursuit-evasion scenarios, significantly enhancing the robustness and adaptability of the formation system. Simulation and physical experiments show that the proposed formation control method not only adapts to complex unknown environments, but also autonomously achieves formation recovery in scenarios involving member detachment and formation disruption.

This paper explores breakthroughs from the perspective of UAV navigation architectures and proposes a UAV autonomous navigation method based on aerial–ground cooperative perception to address the challenge of UAV navigation in GPS-denied and unknown environments. The approach consists of two key components. Firstly, a mobile anchor trilateration and environmental modeling method is developed using a multi-UAV system by integrating the visual sensing capabilities of aerial surveillance UAVs with ultra-wideband technology. It constructs a real-time global 3D environmental model and provides precise positioning information, supporting autonomous planning and target guidance for near-ground UAV navigation. Secondly, based on real-time environmental perception, an improved D* Lite algorithm is employed to plan rapid and collision-free flight trajectories for near-ground navigation. This allows the UAV to autonomously execute collision-free movement from the initial position to the target position in complex environments. The results of real-world flight experiments demonstrate that the system can efficiently construct a global 3D environmental model in real time. It also provides accurate flight trajectories for the near-ground navigation of UAVs while delivering real-time positional updates during flight. The system enables UAVs to autonomously navigate in GPS-denied and unknown environments, and this work verifies the practicality and effectiveness of the proposed air–ground cooperative perception navigation system, as well as the mobile anchor trilateration and environmental modeling method.

With the rapid growth in low-voltage electricity demand, abnormal electricity consumption behavior is becoming more and more frequent, which not only threatens the safe and stable operation of power systems, but also causes huge economic losses. In order to effectively meet this challenge, it is of great practical significance to carry out monitoring and analysis of abnormal power consumption of low-voltage users. In this paper, a new detection model of abnormal power consumption behavior of low-voltage power users in power system based on the hybrid model, namely the K-GBDT model, is proposed. The model combines the GBDT (Gradient Boosting Decision Tree) algorithm with the KNN (K-Nearest Neighbor) algorithm, effectively leveraging the strengths of both approaches. The K-GBDT model employs a two-stage classification strategy. In the first stage, the GBDT algorithm leverages its robust feature learning and nonlinear classification capabilities to perform coarse-grained classification, extracting global patterns and categorical information. In the second stage, based on the coarse classification results from GBDT, the data are partitioned into multiple subsets, and the KNN algorithm is applied to fine classification within each subset. This hybrid approach enables the K-GBDT model to effectively integrate GBDT’s global modeling strength with KNN’s local classification advantages. Comparative experiments and practical applications of the K-GBDT model against standalone GBDT and KNN algorithms were conducted. To further validate the proposed method, a comparative analysis was conducted against the Long Short-Term Memory Autoencoder (LSTM-AE) model. The experimental results demonstrate that the proposed K-GBDT model outperforms single-algorithm models in both classification accuracy and model generalization capability, enabling more accurate identification of abnormal electricity consumption behaviors among low-voltage users. This provides reliable technical support for intelligent management in power systems.

High-precision direction of arrival (DOA) of wideband signals is a very important technology in the field of radar and communication. In this work, we propose an efficient support vector regression (SVR) architecture via a genetic algorithm (GA) for wideband DOA estimation, which exhibits high estimation performance and generalization performance. By adopting the two-sided correlation transformation (TCT) algorithm, the network is trained only from reference frequency data to increase the training efficiency. In order to reduce the redundant information in the array covariance matrix and lower the dimensionality of the input features, the array covariance matrix at a single frequency point is preprocessed according to its conjugate symmetry and elemental characteristics, and the dimensionality-reduced input features are obtained. Specifically, the dimensionality of the input features does not increase with the number of sub-bands when dealing with broadband signals or ultra-broadband signals, which can significantly reduce the training time of the model and the storage capacity of the system. The increased performance of the proposed algorithm is highly desirable in resource-constrained scenarios, and the experimental results demonstrate the efficiency and superiority of the proposed network compared with existing methods.

Aiming at the autonomous navigation of unmanned aerial vehicles (UAVs) in complex and unknown environments, this article combines transfer reinforcement learning theory with an end-to-end mapless visual navigation method and proposes a semantic-driven autonomous navigation method for UAVs. First, we propose a reality reconstruction motion decision model, which consists of two parts: reality reconstruction and motion decision. The reality reconstruction part encodes the environment state into a unified feature space by combining semantic information and raw visual perception to generate a more efficient state description. In the motion decision part, we propose a motion decision network for navigation based on temporal attention and spatial attention mechanisms, which plans its own spatial perception according to key features and important memories to generate highly universal strong temporal navigation motion decisions. Subsequently, the action module generates and executes the optimal actions that the UAV can perform based on the decision results. The simulation results show that the UAV can autonomously explore and actively acquire navigation target position and complete collision-free autonomous navigation from any position to the navigation target in the absence of global information and unknown target positions. The real-world experimental testing results further demonstrate that this method improves the utilization efficiency of perception resources for UAVs, solves the reality gap problem, and has strong transferability. It can be directly transferred to challenging scenarios such as dynamic targets, multiple obstacles, and changes in target and obstacle types without the need for retraining. Overall, it is a highly autonomous and adaptive navigation method.