Xiaoxia Yu

Intelligent following control for electric-drive unmanned ground vehicles (UGVs) is important for human–robot collaboration and autonomous mobility. However, target occlusion, detection noise, and short-term detection failures can reduce following stability and reliability. To address these challenges, this paper proposes a vision-based following method that integrates adaptive Kalman filtering with target-state-aware control. The proposed method updates the observation noise covariance online using an innovation-sequence sliding window and evaluates the validity of visual measurements based on the intersection over union between adjacent frames, thereby enabling target position prediction and compensation under unstable observation conditions. Meanwhile, the longitudinal velocity and lateral angular velocity are dynamically adjusted according to the target detection state, image-center offset, and distance variation to achieve continuous and stable following control. Experimental results show that the proposed method reduces the target position prediction error by approximately 10% compared with conventional Kalman filtering and maintains the longitudinal following distance at about 3 m, improving the robustness, continuity, and motion smoothness of electric-drive UGV target following in complex environments.

Accurate recognition of multistage degradation (MD) in rolling bearings is of great significance for equipment condition monitoring. However, health indicator (HI) constructed using existing single-method approaches often fail to simultaneously achieve degradation sensitivity and monotonicity under varying operating conditions. In addition, mainstream MD recognition methods exhibit limited performance in recognizing transitional samples across degradation stages under cross-condition scenarios. To address these challenges, this study proposes a cross-condition MD recognition method for bearings based on a HI matrix (HIM) and a transition sample recognition enhancement network with multi-branch encoding (TSREN-MBE). Firstly, a degradation sensitive HI (DSHI) is constructed by integrating the least absolute shrinkage and selection operator (LASSO) with the grey wolf optimization algorithm, using comprehensive fault frequency energy as the LASSO regression target, thereby capturing the intrinsic degradation process of bearings. Meanwhile, a time-weighted Wasserstein distance (TWWD) metric is introduced into unsupervised HI construction by incorporating temporal degradation information into the WD, thereby improving the monotonicity of unsupervised HIs. Secondly, DSHI and TWWD are concatenated to form the HIM, which drives the Gath–Geva fuzzy clustering algorithm to adaptively generate MD labels for bearings under different operating conditions. Finally, a TSREN-MBE model is constructed, where multi-branch transformer encoders and multi-head attention are employed to encode and fuse heterogeneous features. By jointly leveraging transition sample recognition enhancement loss, local maximum mean discrepancy, and cross-entropy loss in a joint loss function, the model enhances recognition of transitional samples and improves cross-condition MD recognition accuracy. Experimental results on the XJTU-SY dataset demonstrate the effectiveness and superiority of the proposed method.

In the intelligent detection of welding defects, the sample set is highly imbalanced, with many more normal samples than abnormal ones. However, the existing methods have a weak ability to extract global features in the backbone part, and are prone to favor most types of samples under unbalanced sample conditions. There is still room for further improvement of the existing target detection accuracy. To address these issues, a welding defect detection method is proposed based on a large convolutional inverted-mixed residual network under imbalanced samples. First, UniRepLKNet and SPPF (spatial pyramid pooling and feature fusion) are integrated into the backbone to form the SPPF-UniRepLK module. Second, a large convolutional kernel is introduced in this module to enhance the focus on global features and to capture multi-scale welding defect information better. The inverted-mixed residual mobile block module is designed by combining the inverted residual structure with an attention mechanism, thereby improving the model’s feature extraction for abnormal samples. Third, the Focal Scale-Invariant IoU loss function is developed to adjust category weights dynamically, enhancing the model’s ability to detect abnormal samples. Experimental results demonstrate that the proposed method achieves a precision of 93.2%, a recall of 86.0%, and an mean average precision (mAP) of 89.5% in welding defect detection tasks, surpassing the current state-of-the-art model, GhostNetV2, by margins of 2.2% in precision and 0.9% in mAP, while maintaining comparable recall performance.

Wet clutches are extensively employed in automotive transmission systems due to their benefits of smooth shift and stable operation. However, existing methodologies have not yet thoroughly analyzed the real-time dynamic temperature distribution of wet clutches, and the heating and heat transfer mechanisms during the sliding friction process of friction pairs remain underexplored. To address these gaps, this study proposes a real-time dynamic temperature prediction model for wet clutches and investigates the heat generation and transfer mechanisms in the friction pair sliding process. Specifically, the heat production and exchange dynamics of the wet clutch friction pair are systematically analyzed, followed by an examination of the real-time temperature variation of the separator plate under both high-slip and low-slip speed conditions. In the numerical simulations, the predicted temperature values from the proposed model demonstrate excellent agreement with experimental measurements, with dynamic peak temperature discrepancies remaining within ±2 °C. Furthermore, the validated temperature evolution laws are corroborated by experimental results obtained from a dedicated wet clutch performance test rig, thereby providing comprehensive empirical verification of the proposed real-time dynamic temperature prediction framework for wet clutch separator plates. In summary, the model can accurately capture the temperature variation characteristics of wet clutches under different operating conditions, providing a reliable basis for real-time thermal management of transmission systems. It holds significant practical value for optimizing cooling system design, extending clutch service life, and ensuring shifting quality in vehicles.

To address the challenges of jerky movements and poor tracking performance in outdoor environments for a following‐type mobile robot, a novel marker‐based human–machine following‐motion control strategy is explored. This strategy decouples the control of linear velocity and angular velocity, handling them separately. First, in the design of linear‐velocity control, using the identification of markers to determine the distance between the human and the robot, an enhanced virtual spring model is developed. This involves designing a weighted dynamic damping coefficient to address the rationality issues of the range and trend of the robot's following speed, thereby improving its smoothness and reducing the risk of target loss. Second, in the design of angular velocity control, a new concept of an ‘insensitive zone’ based on the offset of the marker's centre point is proposed, combined with a fuzzy controller to address the issue of robot jitter and enhance its resistance to interference. The experimental results indicate that the average variance in the human–robot distance is 1.037 m, whereas the average variance in the robot's linear velocity is 0.345 m/s. Due to the design of an insensitive region in parameter‐adaptive fuzzy control, the average variance of angular velocity is only 0.031 rad/s. When the human–robot distance exhibits significant fluctuations, the fluctuations in both linear and angular velocities are comparatively small, allowing for stable and smooth following movements. This demonstrates the effectiveness of the motion control strategy designed in this study.

The groove parameters on the friction base of wet clutches significantly affect the temperature distribution of the steel plates. However, existing methods have not thoroughly investigated the mechanisms by which these parameters influence the thermoelastic instability of wet clutches. To address this gap, a comprehensive co-simulation model of the friction sub-multi-physical field was developed to systematically examine the effects of groove inclination, groove density, and groove depth on the surface temperature and mechanical response of the steel plates. The results indicate that both the tilt angle of the grooves and the number of grooves substantially influence the surface temperature distribution of the steel plates. Specifically, increasing the number of grooves leads to a more concentrated distribution of high-temperature hot spots in the circumferential direction, gradually transitioning the surface temperature–hot spot pattern from isolated hot spots to a more uniform high-temperature tropical distribution, which subsequently reduces the maximum surface temperature. On the other hand, increasing the groove inclination angle causes the high-temperature distribution to shift from localized hot spots to a more tropical pattern, with a relatively minor impact on the peak surface temperature. Furthermore, increasing the groove depth results in the dispersion of the high-temperature tropical zone in the circumferential direction, causing the maximum temperature to initially decrease and then increase.

Weld defect detection is crucial for ensuring product reliability and safety in industrial manufacturing. Timely detection of small-target weld defects helps identify potential weld quality issues, minimizing losses and risks in the welding process. Existing methods simply concatenate feature segments during the upsampling process without analyzing the importance of each feature segment, leaving room for further improvement in detection accuracy. To address these issues, this study proposes a feature reorganization network for detecting small target weld defects. First, the C2f-Faster-efficient multi-scale attention feature extraction module and ghost-shuffle convolution are designed, and the large selective kernel network is introduced to dynamically adjust the receptive field of the backbone in the Backbone section, enhancing the model’s ability to extract small target features. Then, a lightweight Content-Aware ReAssembly of FEatures upsampling module is designed in the neck network, which retains more detailed information through feature reorganization and feature expansion, and introduces the parameter-free attention mechanism simple parameter-free attention module to fully capture the contextual information of small targets, thereby enhancing the proposed model’s ability to extract small target features. Finally, the generalized intersection over union boundary loss function is used to improve the network’s bounding box regression performance, achieving intelligent detection of small target weld defects. Experimental results show that the proposed method achieves a mean average precision, parameter count, and computation volume of 85.6%, 2.5 M, and 7.0 G, respectively, for weld defect detection, outperforming the comparison models and meeting the requirements of practical engineering applications.