Zezhou Sun

Dr. Zezhou Sun


Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

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

Representative Publications

View All Publications
16 May 2023 in Symmetry
Verified
Journal Article
Open access
PDF file
JSJunkai Sun
Junkai Sun
ZSZezhou Sun
Zezhou Sun
PWPengfei Wei
Pengfei Wei
BLBin Liu
Bin Liu
YWYaobing Wang
Yaobing Wang
CYChuliang Yan
Chuliang Yan
Aimed at the difficulty of path planning resulting from the variable configuration of the wheel-legged robot for future deep space explorations, this paper proposes a path planning algorithm based on the Theta* algorithm and Timed Elastic Band (TEB) algorithm. Firstly, the structure of the wheel-legged robot is briefly introduced, and the workspace of a single leg is analyzed. Secondly, a method to judge complete obstacles and incomplete obstacles according to the height of the obstacles is proposed alongside a method to search for virtual obstacles, to generate a grid map of the wheel and a grid map of the body, respectively. By dividing obstacles into complete obstacles and incomplete obstacles, the path planning of the wheel-legged robot is split into the planning of the body path and the planning of the wheel path. The body can be still simplified as a point by searching for the virtual obstacle, which avoids the difficulty of a planning path of a variable shape. Then, we proposed hierarchical planning and multiple optimization algorithms for the body path and wheel path based on the Theta* algorithm and TEB algorithm. The path can be optimized and smoothed effectively to obtain a shorter length and higher safety. On that basis, the proposed algorithm is simulated by Matlab. The results of simulations show that the algorithm proposed in this paper can effectively plan the path of the wheel-legged robot by using variable configurations for different types of obstacles. The path-planning algorithm of the wheel-legged robot proposed in this paper has a broad prospect for deep space exploration.

17 April 2023 in Sensors
Verified
Journal Article
Open access
PDF file
JSJunkai Sun
Junkai Sun
ZSZezhou Sun
Zezhou Sun
JLJianfei Li
Jianfei Li
CWChu Wang
Chu Wang
XJXin Jing
Xin Jing
CYChuliang Yan
Chuliang Yan
Addressing the problem that control methods of wheel-legged robots for future Mars exploration missions are too complex, a time-efficient control method based on velocity planning for a hexapod wheel-legged robot is proposed in this paper, which is named time-efficient control based on velocity planning (TeCVP). When the foot end or wheel at knee comes into contact with the ground, the desired velocity of the foot end or knee is transformed according to the velocity transformation of the rigid body from the desired velocity of the torso which is obtained by the deviation of torso position and posture. Furthermore, the torques of joints can be obtained by impedance control. When suspended, the leg is regarded as a system consisting of a virtual spring and a virtual damper to realize control of legs in the swing phase. In addition, leg sequences of switching motion between wheeled configuration and legged configuration are planned. According to a complexity analysis, velocity planning control has lower time complexity and less times of multiplication and addition compared with virtual model control. In addition, simulations show that velocity planning control can realize stable periodic gait motion, wheel-leg switching motion and wheeled motion and the operation time of velocity planning control is about 33.89% less than that of virtual model control, which promises a great prospect for velocity planning control in future planetary exploration missions.

10 November 2023 in Sensors
Verified
Journal Article
Open access
PDF file
LWLihua Wang
Lihua Wang
ZSZezhou Sun
Zezhou Sun
YWYaobing Wang
Yaobing Wang
JWJie Wang
Jie Wang
ZZZhijun Zhao
Zhijun Zhao
CYChuliang Yan
Chuliang Yan
Secure and reliable active debris removal methods are crucial for maintaining the stability of the space environment. Continuum robots, with their hyper-redundant degrees of freedom, offer the ability to capture targets of varying sizes and shapes through whole-arm grasping, making them well-suited for active debris removal missions. This paper proposes a pre-grasping motion planning method for continuum robots based on an improved artificial potential field to restrict the movement area of the grasping target and prevent its escape during the pre-grasping phase. The analysis of the grasping workspace ensures that the target is within the workspace when starting the pre-grasping motion planning by dividing the continuum robot into delivery and grasping segments. An improved artificial potential field is proposed to guide the continuum robot in surrounding the target and creating a grasping area. Specifically, the improved artificial potential field consists of a spatial rotating potential field, an attractive potential field incorporating position and posture potential fields, and a repulsive potential field. The simulation results demonstrate the effectiveness of the proposed method. A comparison of motion planning results between methods that disregard and consider the posture potential field shows that the inclusion of the posture potential field improves the performance of pre-grasping motion planning for spatial targets, achieving a success rate of up to 97.8%.

4 December 2024 in IEEE Access
Verified
Journal Article
Open access
PWPengfei Wei
Pengfei Wei
ZSZezhou Sun
Zezhou Sun
HTHe Tian
He Tian
Rock segmentation on the Martian surface is particularly critical for rover navigation, obstacle avoidance, and scientific target detection. We propose a lightweight bilateral network for semantic segmentation of Martian rock (LBNet). The network consists of a shallow spatial detail branch (SDB) and a deep semantic information branch (SIB). In the shallow spatial detail branch, dense connection channel aggregation convolution (CAConv) is adopted to establish local dependencies for each pixel and preserve detailed information. In the deep semantic information branch, channel split convolution (CSConv) is adopted to extract features by adopting different convolution kernels on different channel, reducing the similarity between different feature maps and increasing feature maps diversity. Finally, a feature fusion module (FFM) is designed to effectively fuse feature maps at different levels. With only 0.37M parameters, the model achieved 93.85% mIoU and 147.8 FPS on the dataset of Perseverance, and 88.62% mIoU and 152.5 FPS on the Curiosity dataset. Experiments show that the model achieves a good balance between segmentation accuracy and inference speed.

15 August 2020
Verified
Book Chapter
Closed access
PDF file
ZSZezhou Sun
Zezhou Sun

15 August 2020
Verified
Book Chapter
Closed access
PDF file
ZSZezhou Sun
Zezhou Sun