Tong Cui

This paper investigates the bipartite synchronization problem in complex networks of agents with three-dimensional nonlinear dynamics. To address the challenges posed by system nonlinearity, heterogeneous node behaviors, and the requirement of fully distributed control, a novel nonlinear saturated controller based on the hyperbolic tangent function is proposed and compared with the traditional linear proportional controller. Furthermore, a unified modeling approach based on the Kronecker product is introduced to compactly represent the global error dynamics, enabling rigorous Lyapunov-based stability analysis and explicit derivation of sufficient conditions for global asymptotic bipartite synchronization. Comparative numerical simulations demonstrate that, compared with the linear controller, the proposed nonlinear controller achieves faster convergence rates, higher synchronization accuracy, and stronger robustness against initial disturbances. The proposed framework is theoretically reliable, practically implementable, and scalable, with potential applications in multi-UAV coordination, distributed sensing, and edge intelligence systems.

In multi-layer multi-pass welding, cumulative molten pool deposition, arc radiation interference, and thermal deformation degrade groove boundary discernibility, rendering conventional machine vision methods inadequate for high-precision seam extraction. To address this, we propose a framework integrating residual network and conditional generative adversarial network integrating laser vision sensing with deep adversarial learning. The model leverages residual network's residual blocks for multi-scale feature retention and conditional generative adversarial network's adversarial training for photorealistic feature map generation. A Color-Attention mechanism dynamically amplifies spectral sensitivity to laser stripes while suppressing arc noise, and an auxiliary centroid loss enforces spatial-geometric consistency between synthesized and actual seams. Experimental results demonstrate that compared to state-of-the-art methods, the proposed method achieves an 80% reduction in Mean Absolute Error(MAE), It also attains a feature extraction rate of 36.72 frames per second (FPS) and keeps error fluctuations below 0.6 pixels under multi-pass superposition conditions. This study establishes a groundbreaking spectral-geometric constrained adversarial framework that comprehensively addresses two critical challenges in multi-layer multi-pass (MLMP) welding: the deterioration of laser stripe contrast due to cumulative spectral interference and the displacement of essential feature points caused by thermal deformation. Unlike previous methodologies, our approach integrates the joint optimization of channel-wise spectral weighting and spatial centroid alignment, thereby achieving unparalleled sub-pixel precision under dynamic industrial conditions.

Failures of Unmanned Aerial Vehicle (UAV) can lead to property damage and casualties, making it imperative to preemptively diagnose failures. Most existing methods do not adequately address the temporal consistency of flight sequence data and often overlook the issue of data imbalance among various fault types. To address these shortcomings, this paper proposes an end-to-end model (CCTFANet) based on the Gated Recurrent Unit (GRU) that ensures temporal consistency of sequence data by maintaining overall causality of the model. Specifically, dual-path causal convolution is used for initial feature extraction from enhanced temporal signals to capture local features at different scales. Subsequently, a temporal feature fusion attention mechanism is introduced as an interactive bridge linking features of different scales, modeling channel dependencies. The GRU is employed for temporal modeling and global feature extraction of fused features to capture long-term dependencies, while maintaining temporal consistency of features in the final stage to ensure the overall causality of the model. This study, utilizing the ALFA dataset, employs time warping and dynamic step sliding window techniques to reconstruct the dataset and alleviate the data imbalance problem. The effectiveness and robustness of the proposed method are validated on the reconstructed dataset. Experimental results demonstrate that the fault diagnosis model can effectively identify UAV faults with a precision of 96.93%. Even under varying noise levels, the proposed method maintains high diagnostic accuracy.

A passive exoskeleton is a wearable robotic device that is worn on the exterior of the user’s body to provide physical support and facilitate movement. Existing elbow passive exoskeletons have limitations in their assistance capabilities and range of applications. In this paper, we propose a controllable elbow passive exoskeleton (CEPE) to address these limitations. The CEPE features a ratchet-based self-energy storage mechanism (RSSM) and a Candan gravity compensation mechanism (CGCM). The CGCM counteracts gravitational forces, while the RSSM stores and releases motion energy. This paper establishes a mathematical model for the RSSM, outlines design specifications for both the RSSM and CEPE, and analyzes the influence of design parameters on the power assistance performance. Three experiments were conducted to validate the feasibility of CEPE, including static strength testing, power assistance without load, and power assistance with load. Results show that, without loading, the CEPE provides compensation effects on elbow joint torque of 68.8%, 93.8%, and 70.7% at shoulder joint angles of 0°, 30°, and 60°, respectively. With a 5 kg loading, adjusting the shoulder joint angle from 30° to 60° results in an increase in the decrease of the average absolute torque directly acting on the elbow joint, from 86% to 91.2%. The adjustable RSSM enables the CEPE to operate in four different modes, expanding its potential applications.

Existing lower limb exoskeletons (LLEs) have demonstrated a lack of sufficient patient involvement during rehabilitation training. To address this issue and better incorporate the patient’s motion intentions, this paper proposes an online brain–computer interface (BCI) system for LLE based motor imagery and stacked ensemble. The establishment of this online BCI system enables a comprehensive closed-loop control process, which includes the collection and decoding of brain signals, robotic control, and real-time feedback mechanisms. Additionally, an online experimental protocol that integrates visual and proprioceptive feedback is developed. To enhance decoding precision, we proposed a novel classification algorithm based on the stacking technique, termed weighted random forests-support vector machines (WRF-SVM). In this algorithm, WRFs function as the base learning models, while SVMs act as the meta-learning layer. To assess the efficacy of the BCI system and the classification algorithm, eight subjects were recruited for testing. The outcomes of both online and offline experiments exhibit high classification accuracy, confirming the viability and utility of the BCI system. We are confident that our approach holds significant promise for practical applications in the field of LLE technology.

Drone vision is widely used in change detection, disaster response, and military reconnaissance due to its wide field of view and flexibility. However, under haze and thin cloud conditions, image quality is usually degraded due to atmospheric scattering. This results in issues like color distortion, reduced contrast, and lower clarity, which negatively impact the performance of subsequent advanced visual tasks. To improve the quality of unmanned aerial vehicle (UAV) images, we propose a dehazing method based on calibration of the atmospheric scattering model. We designed two specialized neural network structures to estimate the two unknown parameters in the atmospheric scattering model: the atmospheric light intensity A and medium transmission t. However, calculation errors always occur in both processes for estimating the two unknown parameters. The error accumulation for atmospheric light and medium transmission will cause the deviation in color fidelity and brightness. Therefore, we designed an encoder-decoder structure for irradiance guidance, which not only eliminates error accumulation but also enhances the detail in the restored image, achieving higher-quality dehazing results. Quantitative and qualitative evaluations indicate that our dehazing method outperforms existing techniques, effectively eliminating haze from drone images and significantly enhancing image clarity and quality in hazy conditions. Specifically, the compared experiment on the R100 dataset demonstrates that the proposed method improved the peak signal-to-noise ratio (PSNR) and structure similarity index measure (SSIM) metrics by 6.9 dB and 0.08 over the second-best method, respectively. On the N100 dataset, the method improved the PSNR and SSIM metrics by 8.7 dB and 0.05 over the second-best method, respectively.

Existing dehazing methods deal with real-world haze images with difficulty, especially scenes with thick haze. One of the main reasons is lacking real-world pair data and robust priors. To improve dehazing ability in real-world scenes, we propose a semi-supervised codebook learning dehazing method. The codebook is used as a strong prior to guide the hazy image recovery process. However, the following two issues arise when the codebook is applied to the image dehazing task: (1) Latent space features obtained from the coding of degraded hazy images suffer from matching errors when nearest-neighbour matching is performed. (2) Maintaining a good balance of image recovery quality and fidelity for heavily degraded dense hazy images is difficult. To reduce the nearest-neighbor matching error rate in the vector quantization stage of VQGAN, we designed the unit dual-attention residual transformer module (UDART) to correct the latent space features. The UDART can make the latent features obtained from the encoding stage closer to those of the corresponding clear image. To balance the quality and fidelity of the dehazing result, we design a haze density guided weight adaptive module (HDGWA), which can adaptively adjust the multi-scale skip connection weights according to haze density. In addition, we use mean teacher, a semi-supervised learning strategy, to bridge the domain gap between synthetic and real-world data and enhance the model generalization in real-world scenes. Comparative experiments show that our method achieves improvements of 0.003, 2.646, and 0.019 over the second-best method for the no-reference metrics FADE, MUSIQ, and DBCNN, respectively, on the real-world dataset URHI.

3D point cloud registration is a critical technology in the fields of visual measurement and robot automation processing. In actual large-scale industrial production, the accuracy of point cloud registration directly affects the quality of automated welding processes. However, most existing methods are confronted with serious challenges such as the failure of partial-to-partial point cloud model registration when facing robot automatic processing guidance and error analysis work. Therefore, this paper proposes a novel two-stage network architecture for point cloud registration, which aims at robot pose adjustment and visual guidance in the field of automated welding by using 3D point cloud data. Specifically, we propose a neighborhood-based multi-head attention module in the coarse registration stage. The neighborhood information of each point can be aggregated through focusing on different weight coefficients of multi-head inputs. Then the spatial structure features which is used to establish the overlapping constraint of point clouds are obtained based on above neighborhood information. In the fine registration stage, we propose the similarity matching removal module based on multiple attention fusion features to explore deeper features from different aspects. By using deep fusion features to guide the similarity calculation, the interference of non-overlapping points is removed to achieve the finer registration. Eventually, we compare and analyze the proposed method with the SOTA ones through several error metrics and overlap estimation experiments based on the ModelNet40 dataset. The test results indicate that our method, relative to other mainstream techniques, achieves lower error rates and the most superior accuracy of 98.61% and recall of 98.37%. To demonstrate the generalization performance of proposed algorithm, extensive experiments on the Stanford 3D Scanning Repository, 7-Scenes and our own scanning dataset using partially overlapping point clouds individually under clean and noisy conditions show the validity and reliability of our proposed registration network.

Mainstream pedestrian recognition algorithms have problems such as low accuracy and insufficient real-time performance. In this study, we developed an improved pedestrian recognition algorithm named YOLO-MSP (multiscale parallel) based on residual network ideas, and we improved the network architecture based on YOLOv5s. Three pooling layers were used in parallel in the MSP module to output multiscale features and improve the accuracy of the model while ensuring real-time performance. The Swin Transformer module was also introduced into the network, which improved the efficiency of the model in image processing by avoiding global calculations. The CBAM (Convolutional Block Attention Module) attention mechanism was added to the C3 module, and this new module was named the CBAMC3 module, which improved model efficiency while ensuring the model was lightweight. The WMD-IOU (weighted multidimensional IOU) loss function proposed in this study used the shape change between the recognition frame and the real frame as a parameter to calculate the loss of the recognition frame shape, which could guide the model to better learn the shape and size of the target and optimize recognition performance. Comparative experiments using the INRIA public data set showed that the proposed YOLO-MSP algorithm outperformed state-of-the-art pedestrian recognition methods in accuracy and speed.