Zengxiang Zhou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Zengxiang Zhou has a total of 78 co-authors. Here are the top co-authors by number of shared publications:
Jianping Wang
University of Science and Technology of China
18 shared publicationsHongzhuan Hu
University of Science and Technology of China
17 shared publicationsZhigang Liu
University of Science and Technology of China
15 shared publicationsJiaru Chu
University of Science and Technology of China
14 shared publicationsYingfu Wang
10 shared publicationsJiahao Zhou
Department of Precision Machinery and Precision Instrumentation, University of Science and Technology of China, Hefei, People’s Republic of China
7 shared publications
Latest Publications
- Open Access
- 1 June 2026
- Published by IOP Publishing in Publications of the Astronomical Society of the Pacific
- Vol. 138 (6) , 064510
Real-time Collision Prediction for High-Density Robotic Fiber Positioners of Multi-Object Spectroscopic Survey Telescope
The growing density of robotic fiber positioners (RFPs) in modern multi-object spectroscopic telescopes, such as the phase II system of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST), makes real-time collision prediction increasingly important, particularly as future instruments are expected to integrate 8000–12,000 RFPs within highly overlapping workspaces. To enhance operational safety and maximize observational efficiency in such high-density configurations, we developed a real-time collision prediction system tailored for astronomical fiber positioning systems. Integrating a field-programmable gate array-based vision system with a lightweight spatiotemporal deep learning model, our method achieves high-precision dynamic RFP trajectory prediction to further forecast potential collisions. The system captures positional variations via optical flow and centroid tracking, enabling reliable motion inference through the deep learning model under realistic observing conditions. Experimental validation using LAMOST observational data shows a 48.19% improvement in trajectory prediction accuracy over the standard Transformer baseline, as measured by final displacement error. Collision-prediction experiments in both simulation and on a real platform further verified the effectiveness and robustness of the proposed framework. On the real platform, our approach achieved 98.98% precision and 99.85% recall. Under batched GPU array-level inference, the total latency remains well below the 6 s prediction horizon. These results demonstrate that our approach can provide accurate and low-latency collision-risk prediction for dense robotic fiber positioning systems, offering a reliable early-warning capability and a practical basis for future collision-avoidance systems in large-scale telescopes equipped with high-density fiber positioners.
- 1 June 2026
- Published by IOP Publishing in Publications of the Astronomical Society of the Pacific
- Vol. 138 (6) , 064510
Open AccessReal-time Collision Prediction for High-Density Robotic Fiber Positioners of Multi-Object Spectroscopic Survey Telescope
The growing density of robotic fiber positioners (RFPs) in modern multi-object spectroscopic telescopes, such as the phase II system of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST), makes real-time collision prediction increasingly important, particularly as future instruments are expected to integrate 8000–12,000 RFPs within highly overlapping workspaces. To enhance operational safety and maximize observational efficiency in such high-density configurations, we developed a real-time collision prediction system tailored for astronomical fiber positioning systems. Integrating a field-programmable gate array-based vision system with a lightweight spatiotemporal deep learning model, our method achieves high-precision dynamic RFP trajectory prediction to further forecast potential collisions. The system captures positional variations via optical flow and centroid tracking, enabling reliable motion inference through the deep learning model under realistic observing conditions. Experimental validation using LAMOST observational data shows a 48.19% improvement in trajectory prediction accuracy over the standard Transformer baseline, as measured by final displacement error. Collision-prediction experiments in both simulation and on a real platform further verified the effectiveness and robustness of the proposed framework. On the real platform, our approach achieved 98.98% precision and 99.85% recall. Under batched GPU array-level inference, the total latency remains well below the 6 s prediction horizon. These results demonstrate that our approach can provide accurate and low-latency collision-risk prediction for dense robotic fiber positioning systems, offering a reliable early-warning capability and a practical basis for future collision-avoidance systems in large-scale telescopes equipped with high-density fiber positioners.
- Open Access
- 13 February 2026
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 171 (3) , 156
Optimizing Structural Parameters and Layout of Fiber Positioners with the Whale Optimization Algorithm for LAMOST
Robotic fiber positioner (RFP) arrays are essential for multiobject spectroscopic telescopes. To improve the survey efficiency, they are densely arranged on the focal planes with overlapped patrol areas, guaranteeing full coverage and increasing the number of RFPs. However, this does not guarantee a proportional increase in observable targets due to collisions. To achieve optimal performance of RFPs across various target distributions, an effective method is necessary for optimizing the parameter settings and array layouts. This study addresses this challenge by applying optimization algorithms to minimize collision risks while maximizing target coverage in different configurations. Initially, the whale optimization algorithm is used to optimize RFPs with mixed arm-length ratios (three distinct ratios). The design is then refined by jointly optimizing arm length, mixed arm ratios, and variable pitch. Simulations based on the Large Sky Area Multi-Object Fiber Spectroscopic Telescope system show that RFPs with equal arm lengths and a fixed pitch of 10.4 mm provide optimal performance for both uniform and positional targets. This method offers a practical solution for designing and arranging RFPs on focal planes, thereby enhancing the performance of fiber positioning systems. Additionally, it can be applied to other major astronomical instruments such as the Dark Energy Spectroscopic Instrument and the Multi-Object Optical and Near-infrared Spectrograph.
- 13 February 2026
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 171 (3) , 156
Open AccessOptimizing Structural Parameters and Layout of Fiber Positioners with the Whale Optimization Algorithm for LAMOST
Robotic fiber positioner (RFP) arrays are essential for multiobject spectroscopic telescopes. To improve the survey efficiency, they are densely arranged on the focal planes with overlapped patrol areas, guaranteeing full coverage and increasing the number of RFPs. However, this does not guarantee a proportional increase in observable targets due to collisions. To achieve optimal performance of RFPs across various target distributions, an effective method is necessary for optimizing the parameter settings and array layouts. This study addresses this challenge by applying optimization algorithms to minimize collision risks while maximizing target coverage in different configurations. Initially, the whale optimization algorithm is used to optimize RFPs with mixed arm-length ratios (three distinct ratios). The design is then refined by jointly optimizing arm length, mixed arm ratios, and variable pitch. Simulations based on the Large Sky Area Multi-Object Fiber Spectroscopic Telescope system show that RFPs with equal arm lengths and a fixed pitch of 10.4 mm provide optimal performance for both uniform and positional targets. This method offers a practical solution for designing and arranging RFPs on focal planes, thereby enhancing the performance of fiber positioning systems. Additionally, it can be applied to other major astronomical instruments such as the Dark Energy Spectroscopic Instrument and the Multi-Object Optical and Near-infrared Spectrograph.
- 2 February 2026
- Published by Elsevier in Measurement
- Vol. 268
Joint adjustment measurement method for multi-laser trackers based on scale bar
- 2 February 2026
- Published by Elsevier in Measurement
- Vol. 268
Joint adjustment measurement method for multi-laser trackers based on scale bar
- Open Access
- 1 January 2026
- Published by SPIE-Intl Soc Optical Eng in Journal of Astronomical Telescopes, Instruments, and Systems
- Vol. 12 (01) , 1-21
Research on an automatic focusing method for the Large Sky Area Multi-Object Fiber Spectroscopic Telescope closed-loop positioning system based on deep learning
The positioning accuracy of optical fibers directly affects the observation efficiency of spectroscopic survey telescopes. The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) adopts an optical fiber closed-loop positioning system based on a photogrammetric measurement system. A photogrammetric measurement system requires accurate identification of the optical spot position to achieve higher optical fiber positioning accuracy. However, the temperature difference within 1 day of the barrel of LAMOST will reach 40°C during the observation period. The change of air refractive index will lead to nonlinear deviation of the light propagation path of the optical spot, which will lead to the dispersion of the optical spot and cause a huge deviation in the position detection of the optical spot. To address this issue, we propose incorporating an automatic focusing system into the optical fiber closed-loop positioning system. Because each LAMOST’s fiber view camera (FVC) covers a field of view with hundreds of 320 μm-diameter optical fibers, fast and efficient focusing control for a large number of tiny optical fiber targets is highly challenging. Our solution implements a passive autofocus architecture combining an efficient image clarity evaluation method with a rapid focus search algorithm. Existing image sharpness evaluation methods are incompatible with FVC raw optical spot images, and traditional hill-climbing-based focusing algorithms require dozens of adjustments, significantly reducing efficiency. Our approach introduces an adaptive method for evaluating high-resolution optical spot image definition using the mean variance of the optical spot region, significantly reducing computational cost while improving threshold adaptability. Furthermore, we developed a focus search algorithm based on a hybrid convolutional neural network and recurrent neural network architecture. The algorithm predicts defocus distance from optical spot state sequence, achieving 95% single-shot focusing accuracy within the depth of field. It has been successfully applied to the LAMOST’s FVCs and also offers new methods for other focusing search methods based on passive autofocus for cameras.
- 1 January 2026
- Published by SPIE-Intl Soc Optical Eng in Journal of Astronomical Telescopes, Instruments, and Systems
- Vol. 12 (01) , 1-21
Open AccessResearch on an automatic focusing method for the Large Sky Area Multi-Object Fiber Spectroscopic Telescope closed-loop positioning system based on deep learning
The positioning accuracy of optical fibers directly affects the observation efficiency of spectroscopic survey telescopes. The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) adopts an optical fiber closed-loop positioning system based on a photogrammetric measurement system. A photogrammetric measurement system requires accurate identification of the optical spot position to achieve higher optical fiber positioning accuracy. However, the temperature difference within 1 day of the barrel of LAMOST will reach 40°C during the observation period. The change of air refractive index will lead to nonlinear deviation of the light propagation path of the optical spot, which will lead to the dispersion of the optical spot and cause a huge deviation in the position detection of the optical spot. To address this issue, we propose incorporating an automatic focusing system into the optical fiber closed-loop positioning system. Because each LAMOST’s fiber view camera (FVC) covers a field of view with hundreds of 320 μm-diameter optical fibers, fast and efficient focusing control for a large number of tiny optical fiber targets is highly challenging. Our solution implements a passive autofocus architecture combining an efficient image clarity evaluation method with a rapid focus search algorithm. Existing image sharpness evaluation methods are incompatible with FVC raw optical spot images, and traditional hill-climbing-based focusing algorithms require dozens of adjustments, significantly reducing efficiency. Our approach introduces an adaptive method for evaluating high-resolution optical spot image definition using the mean variance of the optical spot region, significantly reducing computational cost while improving threshold adaptability. Furthermore, we developed a focus search algorithm based on a hybrid convolutional neural network and recurrent neural network architecture. The algorithm predicts defocus distance from optical spot state sequence, achieving 95% single-shot focusing accuracy within the depth of field. It has been successfully applied to the LAMOST’s FVCs and also offers new methods for other focusing search methods based on passive autofocus for cameras.
- 22 December 2025
- Published by Springer Nature in The International Journal of Advanced Manufacturing Technology
- Vol. 142 (3-4) , 1901-1916
A hand-eye calibration method based on 3D sphere calibrator
- 22 December 2025
- Published by Springer Nature in The International Journal of Advanced Manufacturing Technology
- Vol. 142 (3-4) , 1901-1916
A hand-eye calibration method based on 3D sphere calibrator
- Open Access
- 7 November 2025
- Published by SAGE Publications in Advances in Mechanical Engineering
- Vol. 17 (11)
A 6-DOF industrial robot kinematic inverse solution solving method based on an improved COA optimization algorithm
Six Degrees of Freedom (6-DOF) industrial robot inverse kinematics solutions are constrained by high-dimensional nonlinearity, multi-solution problems, and traditional algorithms’ local optima tendency, affecting operational accuracy and real-time performance. Both traditional analytical methods and standard swarm intelligence algorithms are confronted with numerous challenges. To address these issues, this paper proposes an inverse solution method based on an improved Coati Optimization Algorithm (COA). The Modified Denavit-Hartenberg (MD-H) kinematic model is introduced to solve the joint singularity problem, and an improved COA based on an elite retention strategy and dynamic weights is proposed. The diversity of the population is enhanced through a mutation mechanism to improve convergence efficiency, and dynamic weights are used to balance position and orientation errors. The experimental results show that when performing inverse kinematics calculations on 10 groups of spatial points, this algorithm outperforms the standard algorithms (Whale Optimization Algorithm (WOA), standard COA, Particle Swarm Optimization Algorithm (PSO), and Genetic Algorithm (GA)) in terms of average number of iterations (40.4 times), average position error (0.0021 mm), and average orientation error (8.5 × 10 −5 rad), meeting the accuracy requirements of industrial robot inverse kinematics and providing a new method for robot inverse solution.
- 7 November 2025
- Published by SAGE Publications in Advances in Mechanical Engineering
- Vol. 17 (11)
Open AccessA 6-DOF industrial robot kinematic inverse solution solving method based on an improved COA optimization algorithm
Six Degrees of Freedom (6-DOF) industrial robot inverse kinematics solutions are constrained by high-dimensional nonlinearity, multi-solution problems, and traditional algorithms’ local optima tendency, affecting operational accuracy and real-time performance. Both traditional analytical methods and standard swarm intelligence algorithms are confronted with numerous challenges. To address these issues, this paper proposes an inverse solution method based on an improved Coati Optimization Algorithm (COA). The Modified Denavit-Hartenberg (MD-H) kinematic model is introduced to solve the joint singularity problem, and an improved COA based on an elite retention strategy and dynamic weights is proposed. The diversity of the population is enhanced through a mutation mechanism to improve convergence efficiency, and dynamic weights are used to balance position and orientation errors. The experimental results show that when performing inverse kinematics calculations on 10 groups of spatial points, this algorithm outperforms the standard algorithms (Whale Optimization Algorithm (WOA), standard COA, Particle Swarm Optimization Algorithm (PSO), and Genetic Algorithm (GA)) in terms of average number of iterations (40.4 times), average position error (0.0021 mm), and average orientation error (8.5 × 10 −5 rad), meeting the accuracy requirements of industrial robot inverse kinematics and providing a new method for robot inverse solution.
- Open Access
- 9 July 2025
- Published by MDPI AG in Journal of Eye Movement Research
- Vol. 18 (4) , 28
GMM-HMM-Based Eye Movement Classification for Efficient and Intuitive Dynamic Human–Computer Interaction Systems
Human–computer interaction (HCI) plays a crucial role across various fields, with eye-tracking technology emerging as a key enabler for intuitive and dynamic control in assistive systems like Assistive Robotic Arms (ARAs). By precisely tracking eye movements, this technology allows for more natural user interaction. However, current systems primarily rely on the single gaze-dependent interaction method, which leads to the “Midas Touch” problem. This highlights the need for real-time eye movement classification in dynamic interactions to ensure accurate and efficient control. This paper proposes a novel Gaussian Mixture Model–Hidden Markov Model (GMM-HMM) classification algorithm aimed at overcoming the limitations of traditional methods in dynamic human–robot interactions. By incorporating sum of squared error (SSE)-based feature extraction and hierarchical training, the proposed algorithm achieves a classification accuracy of 94.39%, significantly outperforming existing approaches. Furthermore, it is integrated with a robotic arm system, enabling gaze trajectory-based dynamic path planning, which reduces the average path planning time to 2.97 milliseconds. The experimental results demonstrate the effectiveness of this approach, offering an efficient and intuitive solution for human–robot interaction in dynamic environments. This work provides a robust framework for future assistive robotic systems, improving interaction intuitiveness and efficiency in complex real-world scenarios.
- 9 July 2025
- Published by MDPI AG in Journal of Eye Movement Research
- Vol. 18 (4) , 28
Open AccessGMM-HMM-Based Eye Movement Classification for Efficient and Intuitive Dynamic Human–Computer Interaction Systems
Human–computer interaction (HCI) plays a crucial role across various fields, with eye-tracking technology emerging as a key enabler for intuitive and dynamic control in assistive systems like Assistive Robotic Arms (ARAs). By precisely tracking eye movements, this technology allows for more natural user interaction. However, current systems primarily rely on the single gaze-dependent interaction method, which leads to the “Midas Touch” problem. This highlights the need for real-time eye movement classification in dynamic interactions to ensure accurate and efficient control. This paper proposes a novel Gaussian Mixture Model–Hidden Markov Model (GMM-HMM) classification algorithm aimed at overcoming the limitations of traditional methods in dynamic human–robot interactions. By incorporating sum of squared error (SSE)-based feature extraction and hierarchical training, the proposed algorithm achieves a classification accuracy of 94.39%, significantly outperforming existing approaches. Furthermore, it is integrated with a robotic arm system, enabling gaze trajectory-based dynamic path planning, which reduces the average path planning time to 2.97 milliseconds. The experimental results demonstrate the effectiveness of this approach, offering an efficient and intuitive solution for human–robot interaction in dynamic environments. This work provides a robust framework for future assistive robotic systems, improving interaction intuitiveness and efficiency in complex real-world scenarios.
- Open Access
- 19 June 2025
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 170 (1) , 36
Collision Probability Evaluation of LAMOST Robotic Fiber Positioners in the Design Phase
Multiobject spectroscopic telescopes are crucial for modern astronomy. Among them, the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST), with thousands of robotic fiber positioners (RFPs), is the one with the highest spectra acquisition rates. To guarantee no blind area, RFPs’ working spaces are overlapped to a certain extent, leading to collisions between RFPs and damage to RFPs. Even with algorithms designed for collision avoidance, collisions still occur during operation. Despite a 99.6% success rate of the current adopted path-planning algorithm, 4% of RFPs fail to reach targets. Fail reasons, such as collisions, cause ongoing negative effects across multiple observation rounds, necessitating the replacement of many RFPs. LAMOST is upgrading its fiber positioning system. Considering the above-mentioned problem in the design phase is necessary. To address this challenge, we propose a mathematical model to assess the collision probability. Then, the proposed model is validated by Monte Carlo simulations. During collision probability calculation, we consider factors such as RFP structure, target allocation, motion requirements, and mechanical errors into consideration. Based on this, we employ the genetic algorithm to optimize RFP arrangements with the lowest collision probability and show its function in the design phase. The proposed method is designed for LAMOST, but it is suitable for the future design of newly proposed spectroscopic telescopes with RFPs.
- 19 June 2025
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 170 (1) , 36
Open AccessCollision Probability Evaluation of LAMOST Robotic Fiber Positioners in the Design Phase
Multiobject spectroscopic telescopes are crucial for modern astronomy. Among them, the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST), with thousands of robotic fiber positioners (RFPs), is the one with the highest spectra acquisition rates. To guarantee no blind area, RFPs’ working spaces are overlapped to a certain extent, leading to collisions between RFPs and damage to RFPs. Even with algorithms designed for collision avoidance, collisions still occur during operation. Despite a 99.6% success rate of the current adopted path-planning algorithm, 4% of RFPs fail to reach targets. Fail reasons, such as collisions, cause ongoing negative effects across multiple observation rounds, necessitating the replacement of many RFPs. LAMOST is upgrading its fiber positioning system. Considering the above-mentioned problem in the design phase is necessary. To address this challenge, we propose a mathematical model to assess the collision probability. Then, the proposed model is validated by Monte Carlo simulations. During collision probability calculation, we consider factors such as RFP structure, target allocation, motion requirements, and mechanical errors into consideration. Based on this, we employ the genetic algorithm to optimize RFP arrangements with the lowest collision probability and show its function in the design phase. The proposed method is designed for LAMOST, but it is suitable for the future design of newly proposed spectroscopic telescopes with RFPs.
- 19 June 2025
- Published by American Astronomical Society in The Astronomical Journal
- p. 36
(undefined)
- Open Access
- 14 January 2025
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 169 (2) , 70
Back-illuminated Light-spot-mapping Algorithm for Dual Rotation Robotic Fiber Positioners Based on an Artificial Probability Potential Field Method
Back-illuminated spot mapping is essential for observations with modern spectroscopic survey telescopes. In the fiber position detection system, after the fiber view camera (FVC) detects the positions of the light spots from the robotic fiber positioners (RFPs), these positions must be mapped to the RFPs individually to complete subsequent calculations of their operational data. The back-illuminated images captured by the FVC show thousands of small light spots, each only hundreds of microns in size, randomly distributed on a completely black background, making accurate one-to-one mapping to the actual RFP positions highly challenging due to the lack of additional RFP features. This study constructs an artificial probability potential field to calculate the likelihood of each light spot appearing in the patrol area of an RFP, employing a global probability potential energy optimization method to establish the optimal mapping relationship. The proposed method successfully maps all light spots in the closed-loop fiber positioning task and is ultimately applied to the fiber positioning system of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope, achieving accurate mapping of 4000 RFPs on the focal plane. This algorithm effectively addresses challenges such as multitarget interference, uniform background, and RFP failures, providing a reliable approach for mapping the back-illuminated points of tens of thousands of dual-rotation RFPs without requiring special backlight designs in future spectral survey telescopes.
- 14 January 2025
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 169 (2) , 70
Open AccessBack-illuminated Light-spot-mapping Algorithm for Dual Rotation Robotic Fiber Positioners Based on an Artificial Probability Potential Field Method
Back-illuminated spot mapping is essential for observations with modern spectroscopic survey telescopes. In the fiber position detection system, after the fiber view camera (FVC) detects the positions of the light spots from the robotic fiber positioners (RFPs), these positions must be mapped to the RFPs individually to complete subsequent calculations of their operational data. The back-illuminated images captured by the FVC show thousands of small light spots, each only hundreds of microns in size, randomly distributed on a completely black background, making accurate one-to-one mapping to the actual RFP positions highly challenging due to the lack of additional RFP features. This study constructs an artificial probability potential field to calculate the likelihood of each light spot appearing in the patrol area of an RFP, employing a global probability potential energy optimization method to establish the optimal mapping relationship. The proposed method successfully maps all light spots in the closed-loop fiber positioning task and is ultimately applied to the fiber positioning system of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope, achieving accurate mapping of 4000 RFPs on the focal plane. This algorithm effectively addresses challenges such as multitarget interference, uniform background, and RFP failures, providing a reliable approach for mapping the back-illuminated points of tens of thousands of dual-rotation RFPs without requiring special backlight designs in future spectral survey telescopes.