Zeyuan Huang

Mr. Zeyuan Huang

Beijing University of Posts and Telecommunications, Beijing, China
orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 March 2023 in Actuators
Verified
Journal Article
Open access
PDF file
YSYue Shen
Yue Shen
QJQingxuan Jia
Qingxuan Jia
RWRuiquan Wang
Ruiquan Wang
ZHZeyuan Huang
Zeyuan Huang
GCGang Chen
Gang Chen
Visual servoing is widely used in the peg-in-hole assembly due to the uncertainty of pose. Humans can easily align the peg with the hole according to key visual points/edges. By imitating human behavior, we propose P2HNet, a learning-based neural network that can directly extract desired landmarks for visual servoing. To avoid collecting and annotating a large number of real images for training, we built a virtual assembly scene to generate many synthetic data for transfer learning. A multi-modal peg-in-hole strategy is then introduced to combine image-based search-and-force-based insertion. P2HNet-based visual servoing and spiral search are used to align the peg with the hole from coarse to fine. Force control is then used to complete the insertion. The strategy exploits the flexibility of neural networks and the stability of traditional methods. The effectiveness of the method was experimentally verified in the D-sub connector assembly with sub-millimeter clearance. The results show that the proposed method can achieve a higher success rate and efficiency than the baseline method in the high-precision peg-in-hole assembly.

15 February 2022 in Entropy
Verified
Journal Article
Open access
PDF file
YSYue Shen
Yue Shen
QJQingxuan Jia
Qingxuan Jia
ZHZeyuan Huang
Zeyuan Huang
RWRuiquan Wang
Ruiquan Wang
JFJunting Fei
Junting Fei
GCGang Chen
Gang Chen
Redundant manipulators are widely used in fields such as human-robot collaboration due to their good flexibility. To ensure efficiency and safety, the manipulator is required to avoid obstacles while tracking a desired trajectory in many tasks. Conventional methods for obstacle avoidance of redundant manipulators may encounter joint singularity or exceed joint position limits while tracking the desired trajectory. By integrating deep reinforcement learning into the gradient projection method, a reactive obstacle avoidance method for redundant manipulators is proposed. We establish a general DRL framework for obstacle avoidance, and then a reinforcement learning agent is applied to learn motion in the null space of the redundant manipulator Jacobian matrix. The reward function of reinforcement learning is redesigned to handle multiple constraints automatically. Specifically, the manipulability index is introduced into the reward function, and thus the manipulator can maintain high manipulability to avoid joint singularity while executing tasks. To show the effectiveness of the proposed method, the simulation of 4 degrees of planar manipulator freedom is given. Compared with the gradient projection method, the proposed method outperforms in a success rate of obstacles avoidance, average manipulability, and time efficiency.

8 February 2023 in Actuators
Verified
Journal Article
Open access
PDF file
ZHZeyuan Huang
Zeyuan Huang
GCGang Chen
Gang Chen
YSYue Shen
Yue Shen
RWRuiquan Wang
Ruiquan Wang
CLChuankai Liu
Chuankai Liu
LZLong Zhang
Long Zhang
On-orbit operation tasks require the space robot to work in an unstructured dynamic environment, where the end-effector’s trajectory and obstacle avoidance need to be guaranteed simultaneously. To ensure the completability and safety of the tasks, this paper proposes a new obstacle-avoidance motion planning method for redundant space robots via reinforcement learning (RL). First, the motion planning framework, which combines RL with the null-space motion for redundant space robots, is proposed according to the decomposition of joint motion. Second, the RL model for null-space obstacle avoidance is constructed, where the RL agent’s state and reward function are defined independent of the specific information of obstacles so that it can adapt to dynamic environmental changes. Finally, a curriculum learning-based training strategy for RL agents is designed to improve sample efficiency, training stability, and obstacle-avoidance performance. The simulation shows that the proposed method realizes reactive obstacle avoidance while maintaining the end-effector’s predetermined trajectory, as well as the adaptability to unstructured dynamic environments and robustness to the space robot’s dynamic parameters.

8 March 2022 in Aerospace
Verified
Journal Article
Open access
PDF file
GCGang Chen
Gang Chen
YHyou hong
you hong
ZHZeyuan Huang
Zeyuan Huang
JFJunting Fei
Junting Fei
YWYifan Wang
Yifan Wang
CLChuankai Liu
Chuankai Liu
This paper presents an efficient path planning method for the lunar rover to improve the autonomy and exploration ability in the complex and unstructured lunar surface environment. Firstly, the safe zone for the rover’s motion is defined, based on which a detecting point selection strategy is proposed to choose target positions that meet the rover’s constraints. Secondly, an improved sampling-based path planning method is proposed to get a safe path for the rover efficiently. Thirdly, a map extension strategy for the unstructured and continually varying environment is included to update the roadmap, which takes advantage of the historical planning information. Finally, the proposed method is tested in a complex lunar surface environment. Numerical results show that the appropriate detecting positions can be selected autonomously, while a safe path to the selected detecting position can be obtained with high efficiency and quality compared with the Probabilistic Roadmap (PRM) and A* search algorithm.

7 April 2025 in Applied Intelligence
Verified
Journal Article
Open access
PDF file
JTJiayi Tan
Jiayi Tan
GCGang Chen
Gang Chen
ZHZeyuan Huang
Zeyuan Huang
HLHaofeng Liu
Haofeng Liu
Marcelo H.   Ang Jr.Marcelo H. Ang Jr.
Marcelo H. Ang Jr.
Learning an effective manipulation policy with high efficiency in robotics continues to be a significant challenge. In this paper, we propose E-GAIL, which aims to learn manipulation policies efficiently from a limited set of demonstrations with negative corruption and long-term rewards under the framework of GAIL. Specifically, we propose two techniques: 1) Utilizing both short-term and long-term observations to offer additional rewards for training, accelerating convergence. 2) Incorporating negative actions into generated trajectories for corruption to improve data effectiveness and increase success rates. E-GAIL achieves a 25% improvement in success rates across multiple manipulation tasks, requiring 70% fewer episodes for policy convergence, highlighting its efficiency with limited demonstrations. Our video is available at https://youtu.be/bIDfOjYcY54.

28 July 2022 in SCIENTIA SINICA Technologica
Verified
Journal Article
Closed access
GCGang Chen
Gang Chen
ZHZeyuan Huang
Zeyuan Huang
HYHong You
Hong You
YWYiFan WANG
YiFan WANG
CLChuanKai LIU
ChuanKai LIU
JWJunkui Wang
Junkui Wang
在月面采样任务中,受土壤反作用力、能耗等因素影响,机械臂采样位形会影响其采样性能。为了准确评估采样位形对采样性能的影响,提高采样任务执行的可靠性,提出了机械臂采样位形分析方法与优选策略。首先,分析机械臂与采样器、采样器与铲挖执行机构之间的耦合关系,建立机械臂与采样器的联合动力学模型;其次,分析采样过程中机械臂位形与土壤反作用力、机械臂末端抗偏移能力、关节运动行程以及运动能耗之间的关系,并构建量化指标对采样性能进行评估;进而,根据构建的采样性能评价指标,构建考虑采样点区域、关节轨迹等多种约束的多目标最优化问题框架,设计基于NSGA-2多目标遗传算法的采样位形优选策略;最后,以嫦娥五号采样机械臂为对象进行实验分析。实验结果验证了本文提出的采样位形分析方法与优选策略的有效性,相关结论也为嫦娥五号采样机械臂实际采样指令的生成提供了参考依据。