Zhenzhong LIU

Prof. Dr. Zhenzhong LIU

orcid

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

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

Representative Publications

View All Publications
1 March 2022 in Robotics and Autonomous Systems
Verified
Journal Article
Closed access
YWYan Wang
Yan Wang
QSQiyuan Sun
Qiyuan Sun
ZLZhenzhong LIU
Zhenzhong LIU
LGLin Gu
Lin Gu

12 August 2019 in Applied Sciences
Verified
Journal Article
Open access
PDF file
HZHaixing Zhu
Haixing Zhu
WGWeimin Ge
Weimin Ge
ZLZhenzhong LIU
Zhenzhong LIU
In order to realize the non-destructive intelligent identification of weld surface defects, an intelligent recognition method based on deep learning is proposed, which is mainly formed by convolutional neural network (CNN) and forest random. First, the high-level features are automatically learned through the CNN. Random forest is trained with extracted high-level features to predict the classification results. Secondly, the weld surface defects images are collected and preprocessed by image enhancement and threshold segmentation. A database of weld surface defects is established using pre-processed images. Finally, comparative experiments are performed on the weld surface defects database. The results show that the accuracy of the method combined with CNN and random forest can reach 0.9875, and it also demonstrates the method is effective and practical.

15 March 2021 in Advances in Mechanical Engineering
Verified
Journal Article
Open access
PDF file
HZHaibo Zhou
Haibo Zhou
SMShitai Ma
Shitai Ma
GWGuilian Wang
Guilian Wang
YDYuxin Deng
Yuxin Deng
ZLZhenzhong LIU
Zhenzhong LIU
In order to realize the active and compliant motion of the robot, it is necessary to eliminate the impact caused by processing contact. A hybrid control strategy for grinding and polishing robot is proposed based on adaptive impedance control. Firstly, an electrically driven linear end effector is designed for the robot system. The macro and micro motions control model of the robot is established, by using impedance control method, which based on the contact model of the robot system and the environment. Secondly, the active compliance method is adopted to establish adaptive force control and position tracking control strategies under impact conditions. Finally, the algorithm is verified by Simulink simulation and experiment. The simulation results are as follows: The position tracking error does not exceed 0.009 m, and the steady-state error of the force is less than 1 N. The experimental results show that the motion curve coincides with the surface morphology of the workpiece, and the contact force is stable at 10 ± 3 N. The algorithm can realize more accurate position tracking and force tracking, and provide a reference for the grinding and polishing robot to realize surface processing.

18 September 2019 in Electronics
Verified
Journal Article
Open access
PDF file
HZHaibo Zhou
Haibo Zhou
RCRui Chen
Rui Chen
SZShun Zhou
Shun Zhou
ZLZhenzhong LIU
Zhenzhong LIU
Because the proportional–integral–derivative (PID) parameters selected by experience are random, the control effect of fuzzy PID cannot be optimized. In order to improve the accuracy and stability of robot motion control, an orthogonal-fuzzy PID intelligent control method is proposed. In this paper, the electric steering gear is used as the actuator, and the mathematical model of the servo motor joint drive system is established. The simulation analysis of the original control, PID control, fuzzy PID control, and orthogonal-fuzzy PID control of the manipulator joints in the Simulink software simulation environment and the motion control experiment of the manipulator show that using the orthogonal test method to adjust the PID parameters can quickly determine the appropriate PID parameters and greatly reduce the number of trials. The rise time, adjustment time, and overshoot of the system are significantly reduced by using fuzzy PID control, which can improve the adaptability of the system. By comparing and analyzing fuzzy PID and orthogonal-fuzzy PID control methods, it can be found that the system of orthogonal-fuzzy PID for optimal factor level combination (Kp = 0.1, Ki = 30 and Kd = 0.02) is the optimal system. The experiment results show that the orthogonal-fuzzy PID can further improve the accuracy of the system and reduce the oscillation process of the system near the steady state and make the motion more stable.

31 January 2020 in Applied Sciences
Verified
Journal Article
Open access
PDF file
HZHaibo Zhou
Haibo Zhou
SZShun Zhou
Shun Zhou
JYJia Yu
Jia Yu
ZZZhongdang Zhang
Zhongdang Zhang
ZLZhenzhong LIU
Zhenzhong LIU
In order to realize the technique of quick picking and obstacle avoidance, this work proposes a trajectory optimization method for the pickup manipulator under the obstacle condition. The proposed method is based on the improved artificial potential field method and the cosine adaptive genetic algorithm. Firstly, the Denavit–Hartenberg (D-H) method is used to carry out the kinematics modeling of the pickup manipulator. Taking into account the motion constraints, the cosine adaptive genetic algorithm is utilized to complete the time-optimal trajectory planning. Then, for the collision problem in the obstacle environment, the artificial potential field method is used to establish the attraction, repulsion, and resultant potential field functions. By improving the repulsion potential field function and increasing the sub-target point, obstacle avoidance planning of the improved artificial potential field method is completed. Finally, combined with the improved artificial potential field method and cosine adaptive genetic algorithm, the movement simulation analysis of the five-Degree-of-Freedom pickup manipulator is carried out. The trajectory optimization under the obstacle environment is realized, and the picking efficiency is improved.

10 November 2020 in Frontiers in Neuroinformatics
Verified
Journal Article
Open access
PDF file
LGLin Gu
Lin Gu
XZXiaowei Zhang
Xiaowei Zhang
SYShaodi You
Shaodi You
SZShen Zhao
Shen Zhao
ZLZhenzhong LIU
Zhenzhong LIU
THTatsuya Harada
Tatsuya Harada
One major challenge in medical imaging analysis is the lack of label and annotation which usually requires medical knowledge and training. This issue is particularly serious in the brain image analysis such as the analysis of retinal vasculature, which directly reflects the vascular condition of Central Nervous System (CNS). In this paper, we present a novel semi-supervised learning algorithm to boost the performance of random forest under limited labeled data by exploiting the local structure of unlabeled data. We identify the key bottleneck of random forest to be the information gain calculation and replace it with a graph-embedded entropy which is more reliable for insufficient labeled data scenario. By properly modifying the training process of standard random forest, our algorithm significantly improves the performance while preserving the virtue of random forest such as low computational burden and robustness over over-fitting. Our method has shown a superior performance on both medical imaging analysis and machine learning benchmarks.