The dexterous multi-fingered robotic hand， serving as a key end-effector， is pivotal for enabling robots to perform fine-grained grasping and compliant manipulation. Its advancement holds significant importance for promoting automation in manufacturing， enhancing the intelligence of service robots， and expanding applications in specialized environments. Focusing on humanoid multi-fingered dexterous hand technologies， this paper systematically reviews the current state-of-the-art and future trends. It begins by elucidating the fundamental concepts， system architecture， and typical characteristics of dexterous hands. This is followed by a comprehensive of research achievements from domestic and international teams and commercially available mainstream multi-fingered dexterous hand products， covering various degrees-of-freedom designs and their respective hardware and software implementations. Key technologies， including core hardware components， multi-modal sensory fusion， and control strategies， are critically analyzed. The paper subsequently summarizes practical applications across domains such as industrial assembly， daily life assistance， and operations in extreme environments. Current challenges， particularly in reliability， multi-modal coordination， generalization capability， human-robot safety， and integration and application， are identified. Finally， future research directions are prospected from multiple perspectives， including standard establishment， novel mechanical structures， advanced multi-modal perception and fusion， bionic evolution， and embodied intelligence， aiming to provide valuable insights for in-depth research and groundbreaking applications of dexterous hands.

As an important component of intelligent manufacturing and intelligent operation and maintenance systems， industrial inspection robots are playing a key role in various complex industrial scenarios. With the continuous progress of deep learning， multi-sensor fusion， and autonomous navigation technologies， industrial inspection robots have significantly been improved in terms of accuracy， efficiency， and adaptability. This article systematically reviews the concept， key technologies， and typical applications of industrial inspection robots， and focuses on analyzing the research status of core technologies such as perception and recognition， autonomous positioning and navigation， advanced control， and intelligent decision-making. It also assesses the maturity and industrialization progress of current technologies by combining practical applications in fields such as power， workshops， and special environments. Despite significant achievements in this field， challenges still exist in perception accuracy， dynamic environment adaptability， and task execution intelligence. The development of key technologies is expected to continue in the directions of multi-source data fusion， autonomous learning， and collaborative operation. The article aims to provide a systematic reference and guidance for future research and industrial development of industrial inspection robot technology.

To address the inefficiency of traditional manual visual inventory counting and the high deployment costs of existing automated solutions in Automated Storage and Retrieval Systems （AS/RS）， this paper proposes an intra-warehouse visual inventory system based on modular visual devices. A retrofit-free stacker-accessible modular visual inventory device is designed. On the basis of a YOLOv8-powered visual inventory algorithm for multi-surface information fusion from a single view， the accurate counting of complex stack patterns （e.g.， non-full stacks and staggered stacks） is effectively solved by combining front pallet layer identification with top pallet layer counting. The system also features a non-intrusive integration architecture between the Warehouse Visual Stock System （WVSS） and the existing Warehouse Control System （WCS） via a database， enabling dynamic task scheduling and data closed-loop. Experimental results on four palletized cargo datasets demonstrate a stack quantity recognition accuracy of 96.3% with a processing time of 0.11 seconds per storage location. This solution provides a new engineering path for automated warehousing， characterized by high precision， low deployment cost， and minimal operational disruption.

In recent years， indigenously developed， aerospace-specific 3D structural design systems in China have undergone robust development， with notable achievements in the R&D of core components. However， with the widespread adoption of Large Language Models （LLMs）， establishing an effective interface between 3D structural design and AI-driven methodologies remains a central challenge. Furthermore， existing LLMs lack the capacity for precise reasoning over 3D geometry and complex physical fields， such as aerodynamics， which precludes their direct application in the intelligent design of aircraft structures. Among aerospace structural components， the aircraft wing is critical for generating lift. Its design process is highly complex， heavily reliant on expert experience， and tightly coupled with aerodynamic performance. Consequently， traditional design paradigms are characterized by lengthy iteration cycles and substantial costs. To address this challenge， this paper presents Airfoil-LLM， an intelligent design interface for the 3D modeling of aircraft wings， using the wing as a representative case study. Based on the Transformer architecture， this interface integrates natural language encoding with the decoding of CAD modeling sequences to enable intelligent and automated 3D wing generation. To support model training and validation， we have constructed a large-scale 3D wing design dataset. This comprehensive dataset comprises parameterized 3D CAD models， a wide spectrum of flight conditions from subsonic to supersonic regimes， key aerodynamic performance metrics， and multi-level textual descriptions. Experimental results demonstrate that Airfoil-LLM is capable of deeply comprehending textual descriptions ranging from simple geometric attributes to complex， coupled "geometry-performance" requirements. The system generates 3D models that align closely with the design targets in both geometric shape， achieving a maximum Intersection over Union （IoU） of 0.831， and aerodynamic performance.

To address the issues of suboptimal performance， and high rates of missed detection and false detection in steel surface defect detection technology in industrial production environments， an improved YOLO11 algorithm called GCI-YOLO11 has been proposed. Firstly， in the feature extraction part， the GC-C3k2 module based on the GCNet attention mechanism was designed to enhance the algorithm’s capability to extract contextual feature information from images. Secondly， the CARAFE upsampling algorithm was introduced in the neck part to enable the algorithm to aggregate contextual information within a large receptive field， reducing the loss of feature information during the upsampling process. Finally， Inner-CIoU was used to replace CIoU for loss function optimization， and auxiliary regression box was introduced to improve detection accuracy and model generalization capability. Experimental results show that， GCI-YOLO11 achieved improvements of 2.9% and 2.3% in mAP₅₀ and mAP_50-95 on the NEU-DET dataset， and 1.6% and 0.3% in mAP₅₀ and mAP_50-95 on the GC10-DET dataset， showing better detection performance.

Facing the critical strategic demand for enhancing the resilience and security of industrial and supply chains at the national level， the paper focuses on digital twin technology as a key enabler for driving the digital and intelligent transformation of warehousing and logistics systems. A three-stage evolutionary trajectory is systematically outlined， progressing from static modeling to dynamic synchronization and ultimately to intelligent decision-making. In view of the structural challenges in traditional warehousing and logistics systems such as data silos， lagging equipment maintenance， rigid processes， insufficient flexibility， and lack of holistic optimization， this study constructs a systematic empowerment pathway encompassing five dimensions： “omni-domain data integration—intelligent operations and maintenance reconstruction—process simulation optimization—flexible collaborative scheduling—global decision simulation.” The construction of a unified data foundation enables standardized access and high-quality governance of multi-source heterogeneous data； Deployment of a predictive maintenance platform significantly enhances equipment reliability and system continuity； Application of virtual simulation and dynamic optimization technologies achieves intelligent restructuring of warehouse operations and efficiency multiplication； The construction of an elastic resource scheduling mechanism enhances the adaptability of the system to dynamic demands； By building a full-chain simulation decision-making sandbox， the system is boosted from empirically driven local decisions to data- and model-driven global autonomous optimization. In the practical application within the chemical fiber industry， this technology system has increased production efficiency by 5%， reduced operational costs by 15%， improved equipment utilization by 10%， and shortened fault-handling time by 30%. Looking ahead， digital twin technology will evolve toward “holistic coordination and intelligent symbiosis，” providing critical support for constructing autonomous as well as controllable modern warehousing and logistics systems and cultivating new productive forces.

In order to reduce the monitoring and operating pressure of operators during deep peak shaving， an intelligent desulfurization control system is proposed to address the problems of poor measurement accuracy and large inertia and delay in the controlled objects that prevented long-term stable automatic operations. By using BP neural network， a mapping relationship is constructed between signals such as flue gas flow rate， SO₂ concentration in the raw flue gas and slurry pH to achieve soft measurement of slurry pH value； Replacing conventional PID with variable structure predictive control and combining it with more accurate and reasonable feedforward signals ensures the control effect of the desulfurization system under rapidly changing load and coal quality conditions. Later， utilizing the unit ICS system， the desulfurization intelligent control system is successfully applied to a 650 MW unit. The operation results show that after the system is put into operation， the SO₂ concentration at the outlet is stably controlled within 25 mg/m³， and the deviation between the pH value of the slurry and the set value is kept within 0.2， and there are no significant fluctuations during the variable load and pH meter flushing process. The desulfurization is automatically put into operation for a long time， effectively reducing the operating pressure of the operators.

Aiming at the critical problem that real-time industrial defect detection systems are difficult to balance detection speed， accuracy and computational resource constraints in edge computing environments， a fast lightweight industrial defect detection architecture based on an efficient hybrid state space model is proposed. The architecture designs a C2f_EfficientViM_CGLU fast feature extraction module that deeply integrates the global sequence modelling capability of the visual state space model with the efficient local feature enhancement mechanism of convolutional gated linear units， achieving fast and efficient extraction of complex defect features. The HSM-SSD （Hidden State Mixer based State Space Duality） efficient state space modeling mechanism is introduced to process long sequence dependencies with O（n） linear complexity， significantly improving the fast recognition capability for irregularly shaped and sparsely distributed defects. A Slimneck fast lightweight feature fusion network is constructed through GSConv （Ghost Shuffle Convolution） sparse convolution and VoV-GSCSP （Variance of Variance Ghost Shuffle Cross Stage Partial） efficient feature fusion strategies， achieving significant improvements in inference speed and extreme model compression while ensuring detection accuracy. Comparative experimental results on NEU-DET and APDDD standard datasets show that the proposed network architecture achieves mAP50 of 92.13% on NEU-DET dataset， improving 9.77 percentage points compared to the baseline model YOLOv8n， with only 2.9 M parameters and 7.7 GFLOPs computational complexity， reducing parameters by more than 93% compared to the traditional Faster-RCNN method. The mAP50 on APDDD dataset reaches 89.68%， validating the good generalization performance and fast detection capability of the method. This study provides a theoretical foundation and an efficient and feasible fast detection technical solution for real-time quality control in Industry 4.0 intelligent manufacturing environments.

Industrial robots， with their high flexibility and large working range， have gradually become another important processing equipment besides CNC machine tools in national strategic fields such as aerospace and maritime industry in China. The dynamic characteristics of the robot end are dominated by its joints. To improve the dynamic performance of the robot， it is necessary to start with its weak joints， make improvements and innovations on the basis of the existing series form， and explore new high-stiffness driving methods and robot configurations. A 2-RPR robot for milling large propellers is proposed， which includes a six-axis robot main body and a double electric cylinder branch chain. The translation of the electric cylinder drives the rotation of the robot joint， thereby improving the stiffness of the whole robot. In order to meet the processing space requirements of large propellers， the length parameters of each link of the robot are optimized based on the genetic algorithm to realize the optimization of the working space of the whole robot， so that the working space meets the processing range of a single blade and has the maximum utilization rate.

To address the challenges faced by mobile robots in overcoming obstacles in unstructured environments such as agricultural inspections and disaster rescue， this study proposes a design scheme for a four-wheeled mobile chassis that integrates a rocker-steering suspension with a crank-slider mechanism. First， kinematic and dynamic models of the walking mechanism were established to analyze the influence of key configuration parameters （e.g.， support wheel center distance， hinge distance） on terrain adaptability and load platform posture. A multi-objective optimization method was employed to determine the optimal parameter combination （L_F =200 mm，k₁=0.9）. Second， a three-dimensional virtual prototype was developed by integrating a crank-slider mechanism and symmetric frame design. Dynamic simulations conducted on the RecurDyn platform validated the chassis performance in traversing 18 mm speed bumps and 20 mm semi-cylindrical obstacles， showing pitch angle fluctuations within ±3°and peak torque demand ≤15 N·m. Finally， prototype tests demonstrated that the chassis can stably cross 90 mm speed bump-type obstacles under a 75 kg load， with a linear motion speed of 1.8 m/s and a path deviation of less than 20 mm/5 m. The results indicate that this design significantly enhances the terrain adaptability of mobile robots in unstructured environments， providing a reliable mobile platform for agricultural inspection， logistics， disaster rescue， and similar scenarios.

To overcome the limitations of traditional 2D planar grasping and address the challenge of inaccurate position estimation in existing 6D pose estimation algorithms such as Gen6D， this paper proposes an optimized algorithm， Gen6D-Op. For typical robotic grasping scenarios， the algorithm formulates the position estimation error as a constrained optimization problem based on a collinearity assumption， enabling the precise acquisition of object poses. Building on this high-precision pose， we further design two efficient grasping strategies—Vertical Pose and Planar Projection—to enhance grasping efficiency and stability. Experiments demonstrate that Gen6D-Op significantly improves pose estimation accuracy， reducing the total error by 72.3% to 9.48 mm and achieving a multi-object grasping success rate of 94%. Furthermore， applying the designed grasping strategies effectively reduces the robotic arm's joint angle variation and shortens the grasping time.

High-precision localization of unmanned vehicles in complex environments still faces challenges such as insufficient hardware-software coordination and the lack of standardized integration platforms， which restrict the construction of end-to-end data links and cyber-physical closed-loop interactions. To address these issues， this paper proposes a digital-twin-based localization platform architecture for unmanned vehicles. On the hardware side， a multi-source sensing and communication system is developed on the STM32 platform by integrating Global Navigation Satellite System （GNSS）， Ultra-Wideband （UWB）， Inertial Measurement Unit （IMU）， Bluetooth/Wi-Fi， and Narrowband Internet of Things （NB-IoT） modules. These modules are interconnected via high-speed buses and dedicated interfaces， forming the foundation for heterogeneous data acquisition and transmission. On the software side， the platform leverages the industrial Internet identification and resolution system together with digital twin technology to enable bidirectional real-time interaction and closed-loop control between the physical unmanned vehicle and its virtual counterpart. Core functionalities include data management， spatiotemporal localization， path planning， and geofencing. Experimental results demonstrate that the proposed platform ensures ranging accuracy and stability while enhancing system interactivity and integration， thereby providing a comprehensive hardware-software solution for high-precision unmanned vehicle localization.

Aiming at the problems of insufficient coordinated control and poor adaptability due to the single-arm design of the traditional pork cutting robot， a dual-arm segmentation supple control method based on six-dimensional force sensors is proposed， in which the right arm decomposes the cutting task into the orthogonal subspace of force and position control by constructing a master-slave force-position hybrid control framework， adjusting the cutting force and motion trajectory in real time to ensure the smooth cutting， while the left arm tracks the change of the master arm position and attitude in real time based on the Jacobi matrix mapping and acceleration constraint transfer， to ensure the balance and stability of the two arms when they move together. The simulation results show that the control method has good robustness and flexibility， and is able to accomplish the high-precision pork cutting task.

Currently， wall plastering operations primarily rely on manual labor. Although single-degree-of-freedom rail-guided smoothing plastering equipment has emerged， it still requires human assistance， resulting in low efficiency and high labor costs. To enhance the automation level of wall plastering， a wall-mounted dual-arm plastering robot is designed. The robot adopts a design scheme where dual robotic arms handle spraying and smoothing tasks separately， effectively addressing the issues of uneven spraying thickness and low smoothing efficiency in traditional equipment. The system consists of an autonomous navigation chassis， a concrete mixing tow truck， a dual-arm lifting mechanism， and an adaptive force-controlled smoothing tool. A force-position hybrid control algorithm is proposed， which decouples the force control and position control subspaces through a selection matrix， achieving stable force control and precise position tracking in complex environments. In experimental validation， the robot demonstrates superior flatness and high consistency under various wall conditions， significantly improving construction efficiency and surface quality. Compared to traditional manual methods and single-arm equipment， it respectively enhances operational accuracy and production efficiency by notable margins.

In the production of rubber products， the intensive temperature fluctuations and pressure instability during the mixing and extrusion processes not only adversely affect the mixing quality and the plasticization uniformity of the compound， but also significantly impact equipment energy consumption and production efficiency. To address the issues of large temperature fluctuations and poor pressure stability during rubber compounding and extrusion production， an improved fuzzy PID intelligent cooperative control strategy is proposed. By integrating real-time data acquisition from multiple sensors， including temperature， pressure， rotor speed， and torque sensors， an adaptive parameter tuning mechanism based on process characteristics and a pressure-temperature dynamic coupling compensation model are designed. Through this adaptive adjustment process， precise cooperative control of the mixer rotor and extruder screw is achieved， which deeply analyzes and compensates for the mutual interference between pressure and temperature parameters caused by the strong shear heating of the mixer rotor and the conveying compression process of the extruder screw， thereby overcoming the limitations of single-variable control. Experimental results demonstrate that the improved algorithm reduces the overshoot by 11%， shortens the settling time， and achieves a temperature control accuracy of ±0.8 ℃ in the temperature control process. The interference of pressure fluctuations on temperature is reduced from 35% to 12.7%. This system enhances temperature control precision， reduces pressure fluctuations， and significantly improves product uniformity and production efficiency.

Aiming at the collaborative optimization problem of flexible job shop scheduling and AGV material handling in intelligent factory under multi-variety and small-batch production mode， a joint scheduling method of production and handling based on Dueling Double Deep Q-Network （D3QN） is designed to minimize the maximum completion time， and the conflict-free path planning of AGV is realized by combining the A * algorithm with time window. The design algorithm is compared with a variety of rule scheduling methods， and a variety of different scale examples are designed for experimental verification. The results show that the scheduling performance of D3QN algorithm is better， and it has good optimization effect and generalization ability. At the same time， the influence of the number of AGVs on the scheduling performance is analyzed， and finally it is concluded that it conforms to the law of diminishing marginal benefit.

To address the challenges in object-oriented grasping by robotic arms in unstructured scenarios—such as unknown objects， complex environmental interference， and stringent real-time requirements— this paper proposes a Heatmap-Guided 6-DOF Grasp Detection （HGGD） method that integrates Coordinate Attention （CA） with a Memory-Augmented Neural Network （MANN）， establishing a target-guided grasping model. This approach incorporates a CA module within the pixel encoder. After high-resolution feature map output but before dimensionality reduction， key information is weighted by the CA channel branches， and target regions are located by spatial branches， thereby target perception and location accuracy are enhanced. During feature fusion， MANN is embedded to initialize a memory bank with multi-scale features from the training set. During fusion， read/write heads access historical features and fuse them with current features， updating the memory bank based on confidence thresholds to mitigate information decay and redundancy inherent in traditional fusion. Finally， using the object anchor boxes output by the object detection model， point cloud data undergoes sparsification to reduce computational overhead. The evaluation results based on the GrASPNET-1 billion dataset show that， compared with the benchmark model， the proposed method not only achieves a significant improvement in the crawling success rate index， but also effectively shortens the reasoning time consumption of the target crawling detection task. This solution achieves a coordinated improvement in grasping accuracy and efficiency， providing technical support for the precise grasping of robotic arms in complex real-world scenarios.

Due to the multi-degree-of-freedom， strong coupling， and nonlinear characteristics of wheeled humanoid robots， their motion control， especially whole-body motion control， is highly challenging. To address this issue， a whole-body motion control scheme based on robot kinematics and Model Predictive Control （MPC） is proposed. A simplified kinematic model of the wheeled humanoid robot is established， and a quadratic cost function related to the end-effector's pose is defined. The Relaxed Barrier Functions （RBF） are employed to set the constraints for robots to follow during operation， such as joint position limits， control input constraints， collision avoidance constraints， and tip-over prevention constraints. A simulation experiment for end-effector pose planning is designed， and the results demonstrate that the proposed control scheme not only enables the robot to achieve whole-body motion for end-effector control but also satisfies various constraints.

Difficult-to-cut materials are widely used in the aerospace and aviation industries. These materials have characteristics such as high cutting difficulty， high material cost， and difficult calibration experiments. In the finite element simulation modeling process of difficult-to-cut materials， the setting of material mechanical properties and tool chip friction performance， will significantly affect the prediction accuracy of the simulation model. How to achieve efficient acquisition of mechanical property parameters of difficult-to-cut materials， is of great significance to study the rapid and accurate uncertainty calibration strategy of simulation models. Taking the milling process of Ti₂AlNb intermetallic compound as an example， a finite element simulation model uncertainty calibration method for difficult-to-cut materials under the Bayesian framework is proposed. Firstly， an uncertainty analysis of the model is conducted， and a Bayesian based model uncertainties quantification method is proposed. The uncertainty coefficients are solved using the Markov chain Monte Carlo method. Secondly， finite element modeling， simulation experiment design， and simulation dataset construction for milling process are carried out based on finite element simulation software. A surrogate modeling method based on Gaussian process regression and supporting vector regression is proposed. Finally， Ti₂AlNb milling experimental design is carried out， and the working condition dataset is constructed to quantify the JC constitutive parameters and tool chip friction coefficient within the finite element model. The experimental results show that the uncertainty-calibrated finite element simulation model has significantly improved the prediction accuracy， and the relative error in predicting the cutting force of three-dimensional has decreased from 21.47% before calibration to 12.17%.

To address potential limitations in existing attention mechanisms， such as insufficient learning capability and inadequate focus on critical targets， as well as problems in traditional steel surface defect detection networks like low detection accuracy and single-scale feature extraction， a novel iterative Recurrent Channel Attention （RCA） mechanism is designed. This mechanism independently calculates attention weights along horizontal and vertical directions， enhancing the model's positional awareness of features. By employing a recursive strategy to iteratively refine the fused results and reapplying the generated weights to input features， RCA significantly strengthens the network's ability to capture key feature information within similar targets. This substantially improves detection capability for objects at various locations and scales. The RCA mechanism is integrated into the YOLOv8 network architecture， resulting in an improved steel surface defect detection network based on YOLOv8n. Firstly， Switchable Atrous Convolution （SAConv） is introduced to expand the model's receptive field and enhance its perception and adaptability to multi-scale features， enabling more precise feature extraction. Secondly， an adaptive weighted feature fusion module is incorporated into the Neck section of the network， effectively combining global and local features to strengthen multi-level feature fusion capabilities. Finally， the designed RCA mechanism is implemented， followed by extensive ablation and comparative experiments. The experiments demonstrate that the YOLOv8 model with only RCA achieves an mAP@0.5 of 81.1% on the NEU-DET steel surface defect dataset， representing a 3.4% improvement over the baseline model. On the GC10-DET dataset， it achieves an mAP@0.5 of 63.2%， a 1.7% improvement. The complete YOLO-SCR network reduces computational complexity by 12% compared to the baseline model. It achieves an mAP@0.5 of 84.0% on NEU-DET （a 6.3% improvement） and 64.1% on GC10-DET （a 2.6% improvement）， achieving a better balance between detection accuracy and inference speed.

To address multi-source interference issues， including partial occlusion， perspective distortion， large dimensions， and spatial curvature， in machine vision measurement of multi-specification rebar profiles， a machine vision measurement system tailored for the geometric parameters of rebar is designed and constructed. Firstly， a dual-camera near-far vertical layout resolves the conflict between wide field-of-view and high precision. Secondly， an object plane elevation method combined with dual-camera coupling calibration corrects perspective distortions induced by varying rebar specifications. Thirdly， an image preprocessing framework is developed to identify contour regions and locate sub-pixel coordinates， incorporating Bézier interpolation-based occlusion repair. Fourthly， Euclidean clustering-based feature extraction is adopted for deriving geometric parameters. Finally， back-projection registration and geometric scaling converts pixel dimensions to real-world measurements via Zhang’s calibration. Experimental results demonstrated 0.1 mm accuracy validated through manual measurements and 0.24 s per image processing speed， meeting rapid measurement requirements for six rebar profile parameters.

Aiming at the problems existing in space debris capture under microgravity environment， such as easy fragmentation caused by rigid collisions， low mechanism folding ratio， and poor drive adaptability， this paper proposes a fluid-driven capture mechanism that integrates origami structure and soft robotics technology. This mechanism is designed based on the analysis and optimization of the Youshimura origami structure. It realizes trajectory-controllable folding storage and unfolding wrapping through inner cavity pressure adjustment， and has both high folding ratio （≥1：4） and compliant contact characteristics. To verify the feasibility of the scheme， two types of simulation systems were built： 1） A microgravity contact capture multi-body dynamics simulation based on Abaqus， which analyzes the contact force， wrapping success rate and capture efficiency under different debris shapes； 2） A vision-control integrated system based on CoppeliaSim platform co-simulation， which uses YOLOv5 to realize debris pose recognition and PID control algorithm to generate drive commands， so as to verify the automatic capture accuracy under microgravity disturbance. The simulation results show that the wrapping success rates of the mechanism for cubic debris （55 mm×55 mm×55 mm）， cylindrical debris （Φ50 mm×80 mm） and plate-like debris （400 mm×200 mm×2 mm） are 93%， 80% and 33% respectively， which meets the requirements for compliance and autonomy in microgravity space debris capture.

To enhance the compliance of human-robot interaction and the responsiveness to operational intent in collaborative robots operating within free space， this paper proposes an adaptive impedance control strategy based on operational intent. Current research on adaptive impedance/admittance control suffers from insufficient system stability analysis， reliance on additional biosensors， or excessive algorithmic complexity， thereby limiting its application in practical dynamic interaction scenarios. This paper aims to minimise energy consumption during human-robot interaction. By analysing the impact of operational forces and end-effector velocity on system power， a dynamic mapping relationship between damping parameters and operational intent is derived. An adaptive impedance model is established that adjusts its own damping coefficient based on load force and motion velocity， achieving low damping， high velocity， and high responsiveness during dragging operations， while maintaining high damping for precision during manipulation of delicate components. Trailing experiments demonstrate Pearson correlation coefficients exceeding 0.9789 for force-displacement responses across all directions at the robotic end-effector. Under irregular external forces， motion remains smooth with rapid adaptive impedance parameter adjustments， validating this strategy's effectiveness and robustness within dynamic， uncertain human-robot interaction scenarios.

Aiming at the problems of cumbersome deployment and poor adjustment flexibility associated with QR code navigation， which is widely used in current warehouse robotics， this paper designs a global path planning method based on predefined paths and a secondary positioning and pose adjustment method based on the PL-ICP （Point-to-Line Iterative Closest Point） algorithm. The system utilizes ROS2 （Robot Operating System 2） with 2D LiDAR， IMU， and wheel odometry sensors. The methodology employs the Cartographer laser SLAM algorithm to construct a 2D grid obstacle map and robot localization data files. A self-developed RViz2 plugin is utilized to create navigation map files， upon which the A* algorithm generates global paths. The Navigation2 framework is then implemented for local path planning and motion control. Upon arrival at a designated target point， the robot calculates its pose deviation using laser feature data and corrects its position via a PID control algorithm. Simulation tests conducted on the Gazebo platform demonstrate that， after secondary positioning and pose adjustment， the robot achieves a positional accuracy of ±3 mm in the x and y directions， and a heading angle （θ） accuracy of ±0.1°.

The realization of industrialized intelligent construction for steel structures is contingent upon two prerequisites： first， the development of a fully assembled steel structure system that is inherently efficient for repeated disassembly； second， the development of automated assembly robots to address the issues of low efficiency， low precision， and poor quality associated with on-site manual installation. This paper proposes a solution involving an automated assembly robot for the installation of torsion-shear high-strength bolts in Core-tube type steel column joint. Focusing on the assembly process of M16 torsion-shear high-strength bolts， this study emphasizes the mechanism design and structural reliability analysis of the robot end-effector. A hierarchical control system for bolt-hole assembly， based on machine vision， is designed. Simulation experiments demonstrate that the proposed robotic mechanism satisfies the assembly process requirements and significantly enhances the efficiency， precision， quality， and safety of installing torsion-shear high-strength bolts in core-tube column connections.

A dynamic resource scheduling method based on enhanced double Q-learning （EDQL） is proposed to address the issues of insufficient efficiency in dynamic resource scheduling and difficulty in ensuring differentiated service quality when multiple services coexist in the cloud radio access network （C-RAN） of the fifth generation mobile communication （5G） system. Firstly， establish a virtualized slicing architecture for ultra reliable low latency communication （uRLLC）， enhanced mobile broadband （eMBB）， and massive machine type communication （mMTC）， and achieve adaptive correction of resource weights through a multi priority preemption mechanism and dynamic queue adjustment factors. On this basis， design an EDQL algorithm that integrates competitive network architecture and dynamic reward scaling， and combine Markov decision processes to jointly model the network load， channel state， and queue delay. The experimental results show that the proposed method reduces the forced termination probability of uRLLC services by 82.3%， improves the eMBB service completion rate by 41.2%， increases system resource utilization by 28.5%， and reduces the average queuing delay of mMTC by 76.9%， compared to traditional polling scheduling， static priority， and heuristic rule algorithms. This study provides a new paradigm for 5G multi service resource scheduling through the deep integration of virtualized slicing and reinforcement learning.

In order to address the problems encountered during glass defect detection， such as reliance on manual intervention， strong interferences from complex backgrounds， and difficulties in distinguishing categories， an improved algorithm based on YOLOv8s is proposed. The detection performance is improved by integrating omni-dimensional dynamic convolution （ODConv） and Squeeze-and-Excitation attention mechanism （SEAttention）. This algorithm replaces traditional convolution with ODConv in the backbone feature extraction network of YOLOv8， and enhances the feature capture capability for subtle defects on glass surfaces by dynamically adjusting the overlap and receptive field of convolution kernels； In the feature fusion stage， SEAttention is introduced to enhance effective defect features and suppress background noise by reallocating the weights of channel dimensions. The experimental results showed that the improved YOLOv8s achieved a mAP of 96.8% on a self-made industrial glass defect dataset， an increase of 1.9% compared with that before improvement. The comprehensive mAP of different IOU thresholds increased by 1.8%， and the detection speed met the real-time detection requirements of industry. The performance is comprehensively improved， providing an efficient solution for automated detection and classification of glass surface defects.

With the acceleration of intelligent development in rail transit， classifying and grading onboard data for rail transit vehicles has become a core challenge in achieving data asset management and security control. This paper focuses on the on-board data of EMU （Electric Multiple Unit） trains， presents methods for data classification and grading， and in response to the characteristics of onboard data—such as large scale， strong heterogeneity， high real-time requirements， and significant multi-source correlation—analyzes key technologies and implementation paths. It proposes critical steps for conducting data classification and grading， including clarifying regulatory standards， constructing classification and grading templates， assessment and analysis of data classification and grading， and strengthening post-management. The paper also introduces the approach and application of leveraging large models to enhance onboard data classification and grading， demonstrating the feasibility of its large-scale deployment. This work aims to support further improvements in efficient data utilization and safety management within the rail transit industry.

Simultaneous Localization and Mapping （SLAM） tasks are typically based on the assumption of a static environment and often suffer from severe pose estimation accuracy degradation when disturbed by dynamic objects. This paper proposes a visual SLAM algorithm that integrates dynamic feature values and mask distribution values， with the following key innovations： 1） An environmental assessment mechanism based on the Exponential Weighted Moving Average （EWMA） of dynamic feature values is proposed， which achieves accurate and robust perception of environmental dynamics by integrating historical reprojection errors. 2） A lightweight improved YOLO-seg semantic segmentation method is employed for rapid dynamic object identification and pixel-level mask generation， significantly improving efficiency while maintaining accuracy. 3） A local Bundle Adjustment （BA） optimization strategy based on mask distribution values is designed， which constructs a weighting coefficient by introducing mask compactness and area ratio， effectively mitigating the optimization imbalance caused by feature point removal. 4） An adaptive multi-threaded collaborative system architecture is constructed， enabling on-demand triggering of the dynamic judgment and semantic segmentation threads and optimized resource scheduling. Experiments on the TUM RGB-D dataset demonstrate that in highly dynamic scenarios， the proposed algorithm reduces the absolute trajectory error by up to 94.1%， 38.8%， and 77.3%， compared with ORB-SLAM3， Dyna-SLAM， and YOLO-SLAM， respectively， significantly enhancing the accuracy， robustness， and real-time performance of SLAM in dynamic environments.

To address the problem of the crank-slider mechanism to overturn due to excessive reciprocating frequency of the sliding frame in vibrating membrane bioreactors， the instantaneous center of velocity （ICV） line model was derived through analytical solution research on the crank-slider mechanism. Calculations based on this model reveal that when the crank-connecting rod length ratios are 1⁄5， 1⁄4， and 1⁄3， respectively， the instantaneous center line exhibits the following variation pattern： with the crank length fixed， a longer connecting rod causes the overall ICV line to shift to the right. Except at the midpoint， the connecting rod of ICV line approaches a nearly horizontal straight line. Building upon the ICV line model， the variation patterns of the sliding frame's velocity and acceleration were further derived for the overturning limit frequency of 1.9 Hz and the safe frequency of 1.8 Hz within a 10 minutes interval. Using ADAMS Software to verify the effectiveness of the ICV line model for the crank-slider mechanism， providing crucial design theory references for predicting the stable operation of vibration membrane module mechanisms. Extending the ICV line model， the ICV line of the four-bar linkage mechanism was derived and applied to the design of the human knee joint. The ICV lines at two extreme positions （leg upright and bent） were obtained， which are similar to the J-shaped ICV trajectory of the human knee joint. The proposed ICV line model exhibits excellent alignment with the ideal instantaneous center trajectory of human joints， providing a reference for optimizing designs with human-machine compatibility. The ICV line mode established in this paper holds promising application prospects in the field of biomechanics.

To address the issue of printing quality degradation caused by thermal expansion and resulting pressure fluctuations in the central impression cylinder of satellite-type flexographic printing machines， a double-inlet and double-outlet spiral cooling channel structure was designed to reduce the surface temperature and axial temperature difference of the cylinder. Computational fluid dynamics （CFD） simulations were conducted to investigate the influence of the bidirectional spiral channel configuration on the cylinder's surface temperature distribution. On this basis， an orthogonal experimental design was employed for parameter optimization， aiming to minimize the axial temperature difference on the cylinder surface. Three key structural parameters including spiral pitch， channel width， and channel height were selected as variable factors. The optimal parameter combination was obtained： spiral pitch of 1000 mm， channel width of 200 mm， and channel height of 35 mm， under which the axial temperature difference on the cylinder surface was reduced to 1.02 ℃， representing a 58.5% decrease compared to 2.46 ℃ observed with a conventional unidirectional spiral channel. This study provides valuable guidance for the structural optimization of cooling channels and the design of central impression cylinders in flexographic printing systems.

To address the common challenges in industrial equipment bearing fault diagnosis， including cross-operational condition transfer difficulties and target domain label scarcity， this study proposes an intelligent diagnostic method integrating feature enhancement and domain-adversarial learning. By constructing a Symmetric Dot Pattern feature map based on wavelet transform （WT-SDP）， the original one-dimensional vibration signals are mapped into two-dimensional geometric-semantic feature representations， effectively addressing the limitations of traditional methods in translation invariance and inefficient modeling of long-range dependencies. This approach significantly improves feature separability and noise robustness. Furthermore， a domain-adversarial neural network （DANN） framework is designed， incorporating a gradient reversal layer to achieve multi-scale alignment of feature distributions between source and target domains. This eliminates reliance on target domain labels while mitigating domain shift issues， thereby enhancing model generalization under heterogeneous operating conditions. Experimental validation on cross-domain transfer tasks from the Case Western Reserve University bearing dataset （CWRU） and the Dynamic Diagnostic System testbench （DDS） demonstrates that the proposed method achieves an average diagnostic accuracy of 95% on target domains， representing a 20% improvement over baseline models. This research provides a novel solution for cross-domain fault diagnosis in industrial equipment under complex operating conditions， with the proposed method showcasing significant advantages in domain adaptation efficiency.

In order to solve the problem of task allocation and worker fatigue in hybrid flow shop， a mixed integer programming model considering worker fatigue is proposed. According to the characteristics of the problem， the real number coding method is used to solve the problem of process sequencing and worker assignment， and an improved nutcracker optimization algorithm （INOA） is proposed to solve the problem. Firstly， reverse learning population initialization is used to increase the diversity of the population. Secondly， the parameter factors \mu and \gamma are optimized to improve the search efficiency and convergence speed of the algorithm. Finally， in order to improve the robustness of the algorithm， a population diversity monitoring mechanism is introduced， and Cauchy disturbance is used to avoid stagnation caused by too low population diversity. The INOA and other algorithms are tested on 12 functions. The test results show that the performance of INOA is significantly better than other algorithms. Through the example verification of the connection line assembly line， the results further verify the superiority of the algorithm. The proposed optimization scheme can effectively shorten the completion cycle， prevent excessive fatigue of workers， and promote the rational allocation of tasks.

Against the backdrop of Industry 4.0， the demands for device collaboration and real-time communication continue to rise. Traditional industrial Automated Guided Vehicle （AGV） systems face critical challenges such as inconsistent protocol standards， poor compatibility among heterogeneous devices， and insufficient stability in wide-area communication. To address these issues， this paper proposes a collaborative communication scheme for AGVs based on 5G-VXLAN and OPC UA/DDS protocol integration. The goal is to enable cross-platform interaction and multi-protocol adaptation of AGV control information within an ROS 2 distributed system. By employing VXLAN technology to build a scalable logical Layer 2 overlay network atop a Layer 3 physical infrastructure， the solution effectively overcomes the traditional VLAN's limitation of 4094 identifiers and network scalability bottlenecks， forming a logically unified communication plane. Combined with OPC UA and DDS transmission protocols， AGV PLC data can be integrad into the ROS 2 distributed control system， enabling cross-platform command delivery and status monitoring. Experimental results demonstrate that 5G-VXLAN technology ensures stable and real-time data interaction during AGV mobility， while the integration of OPC UA and DDS enables conversion and transmission of data across different protocols.

This paper addresses vibration-induced machining surface defects and severe tool wear in large overhang-tool boring by proposing a design method for a dual-stage series tuned-mass-damped boring-bar based on a three-degree-of-freedom series dynamics model. The model establishes motion differential equations and amplitude ratio expressions of the system. Following the principle of 3 lowest equally high peaks， MATLAB optimizer is employed to select the optimal parameters for the model under different total mass ratios， and data fitting is adopted to establish the functional relationships between total mass ratio and such key parameters as mass distribution ratio， natual frequency ratio， and damping ratio. The derived relationships guide the design of a dual-stage damper. Comparative harmonic and acceleration response analyses demonstrate significantly lower resonance amplitudes and broader， flatter resonance peaks versus single tuned-mass dampers at identical total mass ratios. The dual-stage design also exhibits faster energy dissipation rates and shorter vibration decay time， providing an effective solution for enhancing vibration resistance in large overhang boring tools.

The refrigerant filling line， as a typical discrete manufacturing system， faces challenges such as complex multi-station signal coordination， long control logic verification cycles， and high on-site commissioning costs. Traditional simulation verification methods are inadequate for achieving a closed-loop verification of both the entire line's control logic and physical behavior. To address these issues， this paper proposes a digital twin-driven virtual commissioning system for flexible refrigerant production. By constructing a three-layer collaborative model comprising "Device-Signal-Control"， high consistency between the virtual model and the physical production line is ensured. Additionally， a real-time closed-loop verification mechanism based on the “PLC-OPC-Twin Model” is designed， forming a dynamic loop between control logic， signal interaction， and device behavior， thus enhancing the observability and diagnostic capability of the debugging process. Experiments show that the digital twin model established based on this virtual commissioning system can expose key control logic defects， signal conflicts， and robot path interference in the design and commissioning process， enabling rapid iteration and early verification of the filling line's control strategy， and providing a reusable technical framework for its flexible deployment.

During the inbound phase of warehouse logistics， it is necessary to inspect， count the incoming goods， and update the inventory records. Traditional inbound inventory counting methods require manual visual inspection and the use of handheld terminal devices to input inventory information. Based on this scenario and aligning with the trend of automation and intelligence transformation in modern logistics， an automated visual stocktaking algorithm for pallet-loads using multi-camera collaboration was designed. Multi-angle images of the pallet-load are captured， and deep learning models are used to detect cartons and the pallet. Specifically， for the front， rear， left， and right side view images， an object detection model is used to calculate the quantity， types， and arrangement of the cartons. For the top view image， an instance segmentation model combined with depth information is used to calculate the number of cartons on the top layer of the pallet-load. The detection results from the five surfaces of the pallet-load are comprehensively processed， and the total number of cartons in the entire pallet-load is calculated through spatial logic reasoning. During the counting process， anomaly detection is performed on the pallet-load based on the positional relationships and type information among cartons， achieving full automation of the entire inventory stocktaking task. An experimental setup was built on a conveyor for testing， and it was found that this stocktaking algorithm achieves an average accuracy of 98%. The calculation process of this algorithm is traceable， solving the problems of traditional manual counting， namely high labor cost and the susceptibility to errors caused by fatigue and distraction. This research， starting from the warehouse inbound process， has realized an automated workflow for pallet-load stocktaking， providing both solutions and theoretical foundations for logistics automation.

Tool groove geometry is the core geometric element regulating cutting performance. The traditional development model relying on empirical trial and error faces challenges such as long development cycles （over 45 days）， lack of multi-physics simulation， and a performance optimization potential of 15%~20%. This paper centers on cutting simulation technology， establishing its theoretical framework and standardized application process. At the theoretical level， it elucidates the finite element method （FEM） stress-strain simulation， computational fluid dynamics （CFD） cutting fluid heat transfer analysis， and parameter calibration and application of the Johnson-Cook dynamic constitutive model. At the modeling level， based on DEFORM-3D and ANSYS Workbench， it constructs three-dimensional models of straight grooves， conventional helical grooves， and optimized helical grooves， quantitatively comparing the effects of groove types on cutting force， temperature， and chip morphology. At the parameter optimization level， it conducts sensitivity analysis on key parameters of helical grooves， clarifying the influence weights and synergistic mechanisms. The results show that，compared with straight groove， the optimized helical groove （with a helix angle of 35°， groove width of 1.6 mm， and groove depth of 2.2 mm） reduces the main cutting force by 21.9% （from 356 N to 278 N） and the maximum temperature by 25.8% （from 752 ℃ to 558 ℃）， with a chip curling radius of ≤3.2 mm and a chip evacuation smoothness score of 92. The non-equidistant helical design reduces the temperature gradient from 480 ℃/mm to 250 ℃/mm， and the thermal stress concentration factor from 2.8 to 1.9. This research provides support for precise design of tool groove geometry， promoting the transformation of cutting simulation technology from "auxiliary verification" to "core of design decision-making".

Aiming at the problem that collaborative robots lack force sensors when performing contact force tracking under unknown environmental stiffness changes， this study proposes a variable impedance control method that integrates virtual force sensing and model predictive control to enhance the robot's force tracking accuracy and adaptability in uncertain environments. Firstly， based on motor current， joint angle and angular velocity signals， a virtual force sensor is constructed via a generalized momentum observer， and the end-effector contact force of the robot is estimated online. Secondly， a model predictive controller is designed， which takes the estimated force and internal robot states as inputs， dynamically adjusts impedance parameters and calculates trajectory corrections in real-time， thereby achieving adaptive impedance regulation during the force tracking process. Experimental results show that the generalized momentum-based disturbance observer can effectively estimate the end-effector contact force， where the force estimation error in key degrees of freedom is less than 1 N， and rapid convergence within 4 seconds is achieved， providing a reliable feedback basis for sensorless force tracking. Effective adaptive force tracking control for different environmental stiffnesses is realized by the proposed method without external force sensors， providing a feasible control solution for compliant operations of robots in uncertain environments.

In response to the problems of multiple parameter settings， high repetitive labor intensity， long programming time， and strong experience dependence in the CNC machining programming of carbon fiber reinforced resin based polymer （CFRP） plate components with simple geometric features and a large quantity， research on feature-based automatic programming technology is carried out. Based on the feature definitions of panel， structural plate， and substrate components， a method based on attribute adjacency graph matching is used to complete the feature recognition of the 3D model. The processing operation generation algorithm based on the production rule and processing element priority map is used to achieve automatic feature process decision-making. The improved genetic algorithm is used to optimize the processing route. Then， the intelligent programming of CNC machining programs is realized. Finally， a prototype system was developed on the NX platform and validated using CFRP board components as test objects. The test results showed that the system improved programming efficiency by more than 22 times compared to traditional manual programming operations， with an accuracy rate of 100%. It has been applied in engineering in aerospace manufacturing enterprises.