MS - recent papers

A visual grasping method for collaborative robots in unstructured scenes based on object detection and grasp pose estimation

Mon, 07 Sep 2026 10:20:24 +0200

A visual grasping method for collaborative robots in unstructured scenes based on object detection and grasp pose estimation
Junxiao Liu, Jun Qian, Yunkai Tan, and Rong Zhou
Mech. Sci. Discuss., https://doi.org/10.5194/ms-2026-165,2026
Preprint under review for MS (discussion: open, 0 comments)
Robotic grasping in complex environments is challenging due to variations in object positions, orientations, and backgrounds. This study proposes a vision-based method for collaborative robots that integrates object detection, target mask generation, and grasp position estimation to improve grasping reliability. Experiments on a robot platform demonstrate effective target-specific grasping in unstructured multi-object scenes, showing its potential for intelligent manufacturing applications.

Two-dimensional point suspension characteristics of a magnetic robot driven by dual electromagnets within a fluid pipe

Tue, 25 Aug 2026 10:20:24 +0200

Two-dimensional point suspension characteristics of a magnetic robot driven by dual electromagnets within a fluid pipe
Zhanxiang Cui, Yonghua Lu, and Yun Zhu
Mech. Sci., 17, 825–838, https://doi.org/10.5194/ms-17-825-2026, 2026
Pipeline robots have demonstrated significant potential in cleaning, inspection, and transportation. By leveraging magnetic-levitation technology, we employ a symmetric magnetic drive system to achieve controlled suspension and locomotion of a magnetic robot within fluid-filled pipelines. The robot exhibits robust resistance to flow disturbances, better positioning accuracy, and stable motion performance.

Research on reconfigurable kinematic analysis and type synthesis of planar prismatic Rubik's cube mechanism

Mon, 17 Aug 2026 10:20:24 +0200

Research on reconfigurable kinematic analysis and type synthesis of planar prismatic Rubik's cube mechanism
Dabao Fan, Daxing Zeng, Yuming Zhao, Haibing Feng, Yunjiao Deng, and Ya Liu
Mech. Sci., 17, 813–823, https://doi.org/10.5194/ms-17-813-2026, 2026
A planar prismatic Rubik's cube mechanism is defined as a Rubik's-cube-inspired mechanism where components only carry out planar motion. Based on a physical prototype, this paper illustrates the composition, topological configuration, and reconfigurable motion process of the mechanism. The type synthesis method proposed in this paper not only creates more novel mechanism configurations but also enriches and improves the type synthesis theories for planar mechanisms.

An effective prediction model for dynamic wear evolution of unmodified and modified helical gear pairs

Wed, 29 Jul 2026 10:20:24 +0200

An effective prediction model for dynamic wear evolution of unmodified and modified helical gear pairs
Bing Yuan, Yuzheng Tan, Songtao Zhao, Jingyi Gong, and Hao Dong
Mech. Sci., 17, 799–811, https://doi.org/10.5194/ms-17-799-2026, 2026
We developed a wear model for helical gears with surface modifications, coupling wear with dynamic contact forces and a dynamic wear coefficient. Unlike traditional constant‑rate models, our approach captures real‑time changes in contact stress. Results show wear has little effect on gear stiffness, but its distribution varies strongly with modification type, which also alters vibration evolution. This work aids selection of surface designs to improve gear reliability and reduce noise.

Optimal design and validation of a heavy-duty wave compensation parallel platform utilizing the NSGA-III algorithm

Tue, 28 Jul 2026 10:20:24 +0200

Optimal design and validation of a heavy-duty wave compensation parallel platform utilizing the NSGA-III algorithm
Tianzhong Huang, Renjie Luo, Lifeng Mao, Jia Wang, Fei Zhou, and Guoxing Zhang
Mech. Sci., 17, 783–798, https://doi.org/10.5194/ms-17-783-2026, 2026
This study proposes a multi-objective optimization approach tailored for a heavy-duty wave compensation parallel platform, which integrates an artificial neural network (ANN) surrogate model with an enhanced non-dominated sorting genetic algorithm III (NSGA-III).

Trajectory-tracking control of the UR10 manipulator based on radial basis function neural network and super-twisting sliding mode

Mon, 27 Jul 2026 10:20:24 +0200

Trajectory-tracking control of the UR10 manipulator based on radial basis function neural network and super-twisting sliding mode
Xiaole Ma, Chenghu Jing, Kun Zhang, Chen Chen, and Yanfeng Wang
Mech. Sci., 17, 759–767, https://doi.org/10.5194/ms-17-759-2026, 2026
We developed an improved control method for UR10 robot arms, which typically face issues with unpredictable forces and friction. By combining an online-learning neural network with a robust sliding-mode controller, our approach enables more accurate trajectory tracking and eliminates jitter.

Mechanical responses of high-speed train derailment due to collision with deformable obstacles

Mon, 27 Jul 2026 10:20:24 +0200

Mechanical responses of high-speed train derailment due to collision with deformable obstacles
Lingxiang Kong, Shuguang Yao, and Dongtao Wang
Mech. Sci., 17, 769–782, https://doi.org/10.5194/ms-17-769-2026, 2026
High-speed train collisions with obstacles can cause derailment. To address this concern, we conducted a study to understand how different collision angles and speeds affect train stability. Our key findings show that as the angle increased from 0 to 45°, the speed at which derailment occurred dropped significantly. These results help identify safety speed limits and guide infrastructure design to reduce collision risks, making high-speed rail travel safer for everyone.

An enhanced neural network-based method for predicting step error in a five-axis finishing machining

Mon, 20 Jul 2026 10:20:24 +0200

An enhanced neural network-based method for predicting step error in a five-axis finishing machining
Te Ye, Wei Liu, Jiaping Zhang, Jiawei Chen, Jingyang Yu, Yuhang Zhao, and Ziyu Zhang
Mech. Sci., 17, 747–758, https://doi.org/10.5194/ms-17-747-2026, 2026
Traditional mathematical methods for calculating machining errors before production are often time-consuming. To solve this problem, we developed a data-driven intelligent method that learns from existing data to realize real-time error prediction. Our method achieves over 99 % prediction accuracy and runs significantly faster than conventional approaches, helping factories manufacture high-precision, high-quality parts more efficiently.

Research on kinematics and fractional order active disturbance rejection control of skating training robot

Tue, 14 Jul 2026 10:20:24 +0200

Research on kinematics and fractional order active disturbance rejection control of skating training robot
Baihang Wang, Xue Zhao, Yihang Gong, Longchao Sun, and Zi Yang
Mech. Sci., 17, 731–746, https://doi.org/10.5194/ms-17-731-2026, 2026
Speed skating training demands precise movement, but traditional coaching relies on subjective experience and existing tools are bulky or inflexible. We created a flexible cable-driven robot with enhanced control, cutting tracking errors significantly to enable precise training and inform smart sports equipment development.

Hybrid nonlinear model predictive and linear quadratic balance control for a two-wheeled self-balancing wheelchair

Mon, 06 Jul 2026 10:20:24 +0200

Hybrid nonlinear model predictive and linear quadratic balance control for a two-wheeled self-balancing wheelchair
Haomin Sun, Xinying Zhang, Yaozhi Gu, Shuai Wang, Weiyue Chen, Shuyue Zhang, Tianyu Xu, Hongliu Yu, and Qiaoling Meng
Mech. Sci., 17, 713–729, https://doi.org/10.5194/ms-17-713-2026, 2026
Self-balancing wheelchairs can help people move in tight spaces, but they must stay upright when starting from a large tilt, carrying different users, or facing sudden pushes. We developed and tested a two-stage control approach that first brings the chair back safely from larger tilts and then keeps it steady near upright. Simulations and prototype tests showed faster recovery, better stability under disturbances, and stronger tolerance to user-weight changes, supporting safer designs.

Multi-objective optimization study on the structure of self-excited oscillating pulse cavitation jet nozzles in submerged environments

Fri, 03 Jul 2026 10:20:24 +0200

Multi-objective optimization study on the structure of self-excited oscillating pulse cavitation jet nozzles in submerged environments
Jinhui Fan, Hao Peng, Ke Liu, Yuan Gao, Liang Zhang, and Changkun He
Mech. Sci., 17, 699–712, https://doi.org/10.5194/ms-17-699-2026, 2026
This study uses a relatively novel optimization system to optimize the structure of a self-excited oscillating nozzle. The study identifies inlet radius, lower-nozzle diameter, and cavity length as key variables (combined impact >87 %); achieves a 77.39 % improvement in overall cleaning efficiency through global optimization; and explains the flow field evolution mechanism.

Research on compliance control strategy of elbow–wrist rehabilitation robot based on information fusion

Fri, 26 Jun 2026 10:20:24 +0200

Research on compliance control strategy of elbow–wrist rehabilitation robot based on information fusion
Hui Bian, Hang Shang, Zihan Li, Yifan Xu, Peixuan Du, Mingzhi Wang, Runyang Liu, and Ze Luo
Mech. Sci., 17, 685–698, https://doi.org/10.5194/ms-17-685-2026, 2026
Aiming to address the problems of the large time delay and low compliance of the system caused by the slow speed of motion intention recognition based on the force signal, this paper integrates the force signal and the surface electromyography (sEMG) signal as the input of the elbow–wrist rehabilitation robot, thereby achieving a good balance between fast response and force compliance in the control system.

Curriculum-learning-driven hierarchical multi-agent deep reinforcement learning for collaborative scheduling in complex supply chain networks

Thu, 25 Jun 2026 10:20:24 +0200

Curriculum-learning-driven hierarchical multi-agent deep reinforcement learning for collaborative scheduling in complex supply chain networks
Jingya Dong, Han Zhao, Suyi Zhao, Yijie Wang, Mengfan Guo, Chunhe Song, and Mingliang Xu
Mech. Sci., 17, 671–684, https://doi.org/10.5194/ms-17-671-2026, 2026
Modern supply chains must coordinate customer orders, factory choices, and delivery routes while conditions change quickly. This study developed a learning-based scheduling method that breaks the whole task into connected decisions and trains them step by step. Tests in simulated networks showed faster learning, shorter completion times, and better performance in unfamiliar settings. The results suggest a practical way to improve coordination in production and delivery systems.

A primary–secondary admittance control strategy for dual-Stewart platforms in confined-space aircraft component alignment

Tue, 02 Jun 2026 10:20:24 +0200

A primary–secondary admittance control strategy for dual-Stewart platforms in confined-space aircraft component alignment
Huijun Yu, Jiahao Lin, Ruimin Tan, Kang Liu, and Pengyuan Zhao
Mech. Sci., 17, 657–670, https://doi.org/10.5194/ms-17-657-2026, 2026
This study proposes a novel dual-robot system controlled by a primary–secondary strategy for accurately aligning large, flexible aircraft parts in tight spaces. The primary robot ensures precise positioning, while the secondary robot actively senses and cancels out harmful internal forces, preventing part deformation. Our simulated results show that this method achieves high alignment accuracy and reduces internal forces by over 90.03 %, offering a safe and efficient automated solution for aircraft assembly.

Design and kinematic analysis of a Miura-oriented origami continuum space manipulator with deployable bending capability

Fri, 29 May 2026 10:20:24 +0200

Design and kinematic analysis of a Miura-oriented origami continuum space manipulator with deployable bending capability
Ruiwei Liu, Manjia Su, Jinhui Zhou, Mengyu Zhong, Kengyi Wang, Hongwei Guo, Chunlong Wang, and Haoyu Yang
Mech. Sci., 17, 645–656, https://doi.org/10.5194/ms-17-645-2026, 2026
We developed an origami-inspired robotic gripper to solve the challenge of capturing irregular space debris. The design combines compact foldability for launch with controlled, flexible bending to adapt to different shapes. We built and tested a prototype, showing it can accurately position itself, securely wrap around various objects, and hold payloads reliably. These results confirm the gripper’s effectiveness, offering a promising new solution for safe space debris capture.

Multi-objective structural optimization of submarine cable pallet based on Kriging-MOPSO

Wed, 27 May 2026 10:20:24 +0200

Multi-objective structural optimization of submarine cable pallet based on Kriging-MOPSO
Yuze Wang, Lijie Zuo, Changfang Zou, Cong Li, Hongliang Zhang, Yi Luo, Yunfei Ding, Zechen Qian, and Yuhe Zou
Mech. Sci., 17, 629–644, https://doi.org/10.5194/ms-17-629-2026, 2026
Although the optimal design of various marine equipment pieces has been widely studied, the optimization of an important device – the submarine cable tray – has long been neglected. To address this issue, this study specifically focused on the structural optimization design of large-scale (400 t class) submarine cable trays.

Adaptive compliant milling control for switch rail edge deburring within an FDAFC-based robotic framework

Wed, 27 May 2026 10:20:24 +0200

Adaptive compliant milling control for switch rail edge deburring within an FDAFC-based robotic framework
Kang Xu, Quang Li, Jianyong Li, Wengang Fan, Zhiwei Wu, and Jiameng Liu
Mech. Sci., 17, 615–628, https://doi.org/10.5194/ms-17-615-2026, 2026
Switch rails are essential to railway turnouts, but machining burrs can cause cracks, shorten service life, and affect safety. This study developed a robot-assisted milling method that adjusts tool motion and cutting force in real time along long, complex rail edges. Simulations and experiments showed reduced force fluctuations, improved force tracking, and good surface quality. The results support safer, more consistent, and automated switch rail deburring.

Comparison of error modeling between inverse kinematics and product of exponentials methods for a 5-DOF hybrid perfusion manipulator

Tue, 26 May 2026 10:20:24 +0200

Comparison of error modeling between inverse kinematics and product of exponentials methods for a 5-DOF hybrid perfusion manipulator
Hui Yang, Long Bai, Zhouxiang Jiang, Xiangyun Li, and Zhongjie Long
Mech. Sci., 17, 575–591, https://doi.org/10.5194/ms-17-575-2026, 2026
The error model of the 5-degree-of-freedom hybrid perfusion manipulator established by the product of exponentials formula is more comprehensive and effective. Compared with the traditional inverse kinematic method, the mean pose errors of the moving platform after kinematic calibration based on the product of the exponentials method are reduced by 89.04 % and 63.79 %, respectively. The proposed modeling method is applicable to error modeling and analysis of most parallel mechanisms.

Effect of surface roughness on erosion wear of turboshaft engine compressor blades

Tue, 26 May 2026 10:20:24 +0200

Effect of surface roughness on erosion wear of turboshaft engine compressor blades
Dunyuan Luo, Guangfu Bin, Andrew Ball, Fengshou Gu, Haiyan Miao, Chao Li, Ahmed Hamood, and Wei Yuan
Mech. Sci., 17, 593–613, https://doi.org/10.5194/ms-17-593-2026, 2026
Surface roughness has the greatest impact on the erosion wear of the rotor blade, followed by the stator blade, with the least on the guide blade. As surface roughness increases, the maximum wear rate rises and the concentrated erosion wear area expands, while its location does not shift significantly. The results can provide a basis for the erosion wear assessment of compressor blades at different service stages.

Reliability-aware physics-guided enhancement for sparse-path Lamb-wave defect imaging

Fri, 22 May 2026 10:20:24 +0200

Reliability-aware physics-guided enhancement for sparse-path Lamb-wave defect imaging
Peijiang Li and Ting You
Mech. Sci., 17, 557–573, https://doi.org/10.5194/ms-17-557-2026, 2026
We studied how to locate damage in metal plates when only a few sensor measurements are available. Our method first combines the limited signals in a more reliable way and then sharpens the damage image with a guided learning step. Tests using laboratory measurements and supporting simulated data showed clearer images and more accurate locations. This could make structural inspection more reliable when data are scarce.