It often takes a lot of trial and error to get a quadruped robot to learn a proper and natural gait directly through reinforcement learning. Moreover, it requires plenty of attempts and clever reward settings to learn appropriate locomotion. However, the success rate of network convergence is still relatively low. In this paper, the referred trajectory, inverse kinematics, and transformation loss are integrated into the training process of reinforcement learning as prior knowledge. Therefore reinforcement learning only needs to search for the optimal solution around the referred trajectory, making it easier to find the appropriate locomotion and guarantee convergence. When testing, a PD controller is fused into the trained model to reduce the velocity following error. Based on the above ideas, we propose two control framework - single closed-loop and double closed-loop. And their effectiveness is proved through experiments. It can efficiently help quadruped robots learn appropriate gait and realize smooth and omnidirectional locomotion, which all learned in one model.
{"title":"Learning Smooth and Omnidirectional Locomotion for Quadruped Robots","authors":"Jiaxi Wu, Chenan Wang, Dianmin Zhang, Shanlin Zhong, Boxing Wang, Hong Qiao","doi":"10.1109/ICARM52023.2021.9536204","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536204","url":null,"abstract":"It often takes a lot of trial and error to get a quadruped robot to learn a proper and natural gait directly through reinforcement learning. Moreover, it requires plenty of attempts and clever reward settings to learn appropriate locomotion. However, the success rate of network convergence is still relatively low. In this paper, the referred trajectory, inverse kinematics, and transformation loss are integrated into the training process of reinforcement learning as prior knowledge. Therefore reinforcement learning only needs to search for the optimal solution around the referred trajectory, making it easier to find the appropriate locomotion and guarantee convergence. When testing, a PD controller is fused into the trained model to reduce the velocity following error. Based on the above ideas, we propose two control framework - single closed-loop and double closed-loop. And their effectiveness is proved through experiments. It can efficiently help quadruped robots learn appropriate gait and realize smooth and omnidirectional locomotion, which all learned in one model.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"151 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133864741","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536201
Yu Zhang, Yifan Wu, Linghuan Kong, Yinsong Ma, W. He
In this paper, we proposed an adaptive control scheme for a robotic manipulator with continuous repetitive deferred and constant (CRDC) output performance constraints. A new shifting function for performance errors is introduced and entrapped into barrier Lyapunov function (BLF) to address the negative aspects of barrier functions. By adopting this error shifting function into control synthesis, a tracking control approach considering uncertain initial conditions and external perturbations is first developed for the robotic manipulator to guarantee CRDC output constraints. In the other existing literatures, the system states must satisfy the prescribed constraints initially, and cannot violate the constraints during system operation. However, the novel scheme is able to address the aforementioned situation, and the prescribed constraints can be violated both procedurally and initially. Thus, the proposed method is more applicable. The effectiveness of this novel control scheme is demonstrated in simulation.
{"title":"Tracking Control for a Robotic Manipulator under Constraint Violation during Operation and Unknown Initial Conditions","authors":"Yu Zhang, Yifan Wu, Linghuan Kong, Yinsong Ma, W. He","doi":"10.1109/ICARM52023.2021.9536201","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536201","url":null,"abstract":"In this paper, we proposed an adaptive control scheme for a robotic manipulator with continuous repetitive deferred and constant (CRDC) output performance constraints. A new shifting function for performance errors is introduced and entrapped into barrier Lyapunov function (BLF) to address the negative aspects of barrier functions. By adopting this error shifting function into control synthesis, a tracking control approach considering uncertain initial conditions and external perturbations is first developed for the robotic manipulator to guarantee CRDC output constraints. In the other existing literatures, the system states must satisfy the prescribed constraints initially, and cannot violate the constraints during system operation. However, the novel scheme is able to address the aforementioned situation, and the prescribed constraints can be violated both procedurally and initially. Thus, the proposed method is more applicable. The effectiveness of this novel control scheme is demonstrated in simulation.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115352364","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536202
Xiangyang Zhou, Tongtong Shu, Hao Gao
Non-linear friction disturbance is an important interference factor in the high-precision control of pan-tilt for light and small multirotor unmanned aerial vehicle (MUAV) application. Non-linear friction disturbance compensation is a key to realize high-precision remote sensing imaging. The model-based friction compensation method is difficult to estimate and compensate the non-linear friction disturbance accurately, a non-model friction compensation method based on fuzzy control of a pan-tilt for aerial remote sensing application is presented. In the pan-tilt control system, the fuzzy controller is applied to the position loop to compensate the angular position deviation caused by non-linear friction disturbance. Simulations and experiments are carried out to validate the effectiveness of the proposed method. The results show that, compared with PID control, the fuzzy controller has better friction compensation effect, and the control performance of pan-tilt is improved significantly.
{"title":"Non-model Friction Disturbance Compensation of a Pan-tilt Based on MUAV for Aerial Remote Sensing Application","authors":"Xiangyang Zhou, Tongtong Shu, Hao Gao","doi":"10.1109/ICARM52023.2021.9536202","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536202","url":null,"abstract":"Non-linear friction disturbance is an important interference factor in the high-precision control of pan-tilt for light and small multirotor unmanned aerial vehicle (MUAV) application. Non-linear friction disturbance compensation is a key to realize high-precision remote sensing imaging. The model-based friction compensation method is difficult to estimate and compensate the non-linear friction disturbance accurately, a non-model friction compensation method based on fuzzy control of a pan-tilt for aerial remote sensing application is presented. In the pan-tilt control system, the fuzzy controller is applied to the position loop to compensate the angular position deviation caused by non-linear friction disturbance. Simulations and experiments are carried out to validate the effectiveness of the proposed method. The results show that, compared with PID control, the fuzzy controller has better friction compensation effect, and the control performance of pan-tilt is improved significantly.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"68 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114002739","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536157
X. Zhang, Yansheng Gong, Zhiwei Li, Xuan Liu, Shuyue Pan, Jun Li
Multimodal data fusion is becoming a trend for the field of autonomous driving, especially for lane detection. In the process of driving, sensors often encounter problems such as modality imbalance, changing illumination and so on. Therefore, it is worthwhile to study the problems of applying multimodal fusion for lane detection and modality imbalance in the fusion process. In this paper, we propose a novel multimodal model for lane detection, in which attention mechanism is embedded into network to balance multimodal feature fusion and to improve detection capability. In addition, we use multi-frame input and long short-term memory (LSTM) network to solve the shadow interference, vehicles occlusion and mark degradation. At the same time, the network can be applied to the task of lane detection. In order to verify the effect of multimodal application and attention mechanism on fusion, we have designed adequate experiments on processed continuous scene KITTI dataset. The results show that precision increases by about 15% when LiDAR is added compared with RGB only. Besides, attention mechanism obviously improves the performance of multi-modal detection by balancing multi-modal features.
{"title":"Multi-Modal Attention Guided Real-Time Lane Detection","authors":"X. Zhang, Yansheng Gong, Zhiwei Li, Xuan Liu, Shuyue Pan, Jun Li","doi":"10.1109/ICARM52023.2021.9536157","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536157","url":null,"abstract":"Multimodal data fusion is becoming a trend for the field of autonomous driving, especially for lane detection. In the process of driving, sensors often encounter problems such as modality imbalance, changing illumination and so on. Therefore, it is worthwhile to study the problems of applying multimodal fusion for lane detection and modality imbalance in the fusion process. In this paper, we propose a novel multimodal model for lane detection, in which attention mechanism is embedded into network to balance multimodal feature fusion and to improve detection capability. In addition, we use multi-frame input and long short-term memory (LSTM) network to solve the shadow interference, vehicles occlusion and mark degradation. At the same time, the network can be applied to the task of lane detection. In order to verify the effect of multimodal application and attention mechanism on fusion, we have designed adequate experiments on processed continuous scene KITTI dataset. The results show that precision increases by about 15% when LiDAR is added compared with RGB only. Besides, attention mechanism obviously improves the performance of multi-modal detection by balancing multi-modal features.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"56 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114021854","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536199
Fengxu Wang, Jianqing Peng, Han Yuan, Bin Liang, Wenfu Xu
A cable-driven hyper-redundant manipulator receives great attention due to its slender body and flexible movement. The configuration of the driving cable is the key to manipulator design. At present, cables are arranged at equal angles on the disk, but it lacks theoretical support. The analysis about the influence of cable positions on cable tensions is relatively deficient. In this paper, the force on links is studied. The influence of cable positions on driving force is analyzed. The relationship between cable position, driving force, external force and joint angles are studied. Numerical simulation is carried out to verify the theoretical analysis. The results show the maximum of driving forces is the smallest when cables are arranged at equal angles and decreases as the cable position becomes uniform. The results explain the rationality of the equal configuration of the cables in the cable-driven hyper-redundant manipulator. It’s beneficial to guide the design of cable-driven manipulator.
{"title":"Cable Configuration and Driving Force Analysis of a Cable-Driven Hyper-Redundant Manipulator","authors":"Fengxu Wang, Jianqing Peng, Han Yuan, Bin Liang, Wenfu Xu","doi":"10.1109/ICARM52023.2021.9536199","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536199","url":null,"abstract":"A cable-driven hyper-redundant manipulator receives great attention due to its slender body and flexible movement. The configuration of the driving cable is the key to manipulator design. At present, cables are arranged at equal angles on the disk, but it lacks theoretical support. The analysis about the influence of cable positions on cable tensions is relatively deficient. In this paper, the force on links is studied. The influence of cable positions on driving force is analyzed. The relationship between cable position, driving force, external force and joint angles are studied. Numerical simulation is carried out to verify the theoretical analysis. The results show the maximum of driving forces is the smallest when cables are arranged at equal angles and decreases as the cable position becomes uniform. The results explain the rationality of the equal configuration of the cables in the cable-driven hyper-redundant manipulator. It’s beneficial to guide the design of cable-driven manipulator.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"19 5","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120895551","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536107
Yujun Liu, Yutian Wang, Zeyan Zhuang, Xian Guo
Attitude control of Hypersonic Aircraft is a very challenging subject due to the uncertainties and various noises of the system. In this paper, we propose a new methodology to solve this problem. Firstly, the attitude control of Hypersonic Aircraft is reformulated as a system of Forward-Backward Stochastic Differential Equations. Deep Neural Networks (DNNs) are used to get optimal solution of the equations. We have studied several deep neural networks, including FC-based architecture and LSTM-based architecture and proposed a new FC-based architecture that shares the weights between different time steps, which performed satisfactorily in this problem. The performance and universality of the algorithm are tested in both unconstrained and control-constrained cases. Simulation and experimental results verify the superiority of the algorithm.
{"title":"Deep FBSDE Controller for Attitude Control of Hypersonic Aircraft","authors":"Yujun Liu, Yutian Wang, Zeyan Zhuang, Xian Guo","doi":"10.1109/ICARM52023.2021.9536107","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536107","url":null,"abstract":"Attitude control of Hypersonic Aircraft is a very challenging subject due to the uncertainties and various noises of the system. In this paper, we propose a new methodology to solve this problem. Firstly, the attitude control of Hypersonic Aircraft is reformulated as a system of Forward-Backward Stochastic Differential Equations. Deep Neural Networks (DNNs) are used to get optimal solution of the equations. We have studied several deep neural networks, including FC-based architecture and LSTM-based architecture and proposed a new FC-based architecture that shares the weights between different time steps, which performed satisfactorily in this problem. The performance and universality of the algorithm are tested in both unconstrained and control-constrained cases. Simulation and experimental results verify the superiority of the algorithm.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121378543","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536179
Renpeng Wang, W. Xi, Xian Guo, Yongchun Fang
This paper presents a method for snake robots with orthogonal joints to follow a path via crawler gait. Considering snake robot system’s redundancy, a novel path integral reinforcement learning (PI2) framework is applied to solve it. Taking advantage of crawler gait, the path following problem is first simplified to solve optimal curvature sequence for it. Then rolling optimization algorithm is adopted through the solving process to improve solution efficiency and real-time performance. Moreover, path integral is integrated into the rolling optimization to improve solution quality. Finally, we validate the frame by simulation, with results that follow the target path.
{"title":"Path Following for Snake Robot Using Crawler Gait Based on Path Integral Reinforcement Learning","authors":"Renpeng Wang, W. Xi, Xian Guo, Yongchun Fang","doi":"10.1109/ICARM52023.2021.9536179","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536179","url":null,"abstract":"This paper presents a method for snake robots with orthogonal joints to follow a path via crawler gait. Considering snake robot system’s redundancy, a novel path integral reinforcement learning (PI2) framework is applied to solve it. Taking advantage of crawler gait, the path following problem is first simplified to solve optimal curvature sequence for it. Then rolling optimization algorithm is adopted through the solving process to improve solution efficiency and real-time performance. Moreover, path integral is integrated into the rolling optimization to improve solution quality. Finally, we validate the frame by simulation, with results that follow the target path.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116360459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536166
Xinxing Chen, Kuangen Zhang, Yuquan Leng, Chenglong Fu
Wearable robotic systems have been widely studied in recent years, but it still remains a challenge to design a user-adaptive controller for wearable robotic systems to ensure personalized and accurate human-robot interaction. Accurate human motion classification and person identification are two premises helping design user-adaptive controllers for wearable robotic systems. In this paper, we proposed a multi-task learning method for human motion classification and person identification with a single neural network, which can serve as a solution to personalized human-robot interaction, and can also serve as a benchmark for the following studies in related fields. The multi-task learning neural network was trained and tested on a public human motion data set. The proposed method was capable to classify human motions and identify the person, with 99.13% and 96.51% accuracy, respectively. We also compared the proposed method with a benchmark single task learning method for human motion classification, the results showed that the performance of the multi-task learning method is more superior.
{"title":"A Multi-task Learning Method for Human Motion Classification and Person Identification","authors":"Xinxing Chen, Kuangen Zhang, Yuquan Leng, Chenglong Fu","doi":"10.1109/ICARM52023.2021.9536166","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536166","url":null,"abstract":"Wearable robotic systems have been widely studied in recent years, but it still remains a challenge to design a user-adaptive controller for wearable robotic systems to ensure personalized and accurate human-robot interaction. Accurate human motion classification and person identification are two premises helping design user-adaptive controllers for wearable robotic systems. In this paper, we proposed a multi-task learning method for human motion classification and person identification with a single neural network, which can serve as a solution to personalized human-robot interaction, and can also serve as a benchmark for the following studies in related fields. The multi-task learning neural network was trained and tested on a public human motion data set. The proposed method was capable to classify human motions and identify the person, with 99.13% and 96.51% accuracy, respectively. We also compared the proposed method with a benchmark single task learning method for human motion classification, the results showed that the performance of the multi-task learning method is more superior.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122472567","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536164
Xuehao Sun, Shuchao Deng, Baohong Tong
This paper proposes a novel trajectory planning approach based on time elastic band to solve the problem of dynamic obstacle avoidance of mobile robot. Uncertain factors in the scenario need to be considered in trajectory planning. Thus, this approach includes multiple constraints, such as robot motion speed, motion state, and obstacles. First, to solve the optimal speed of the mobile robot, the workspace potential field must be established, and environmental information should be obtained to constrain the robot speed. Second, a costmap needs to be established to detect dynamic obstacles, and obstacle avoidance strategies based on the relative motion relationship between dynamic obstacles and the robot should be proposed to realize dynamic obstacle avoidance. Finally, by combining multiple constraints, the collision-free trajectory planning from the start point to the target point is completed, and the mobile robot realizes collision-free smooth motion. Experimental results show that this approach has satisfactory obstacle avoidance planning effects and superior kinematics characteristics and improves the comfort and safety of the mobile robot.
{"title":"Trajectory Planning Approach of Mobile Robot Dynamic Obstacle Avoidance with Multiple Constraints","authors":"Xuehao Sun, Shuchao Deng, Baohong Tong","doi":"10.1109/ICARM52023.2021.9536164","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536164","url":null,"abstract":"This paper proposes a novel trajectory planning approach based on time elastic band to solve the problem of dynamic obstacle avoidance of mobile robot. Uncertain factors in the scenario need to be considered in trajectory planning. Thus, this approach includes multiple constraints, such as robot motion speed, motion state, and obstacles. First, to solve the optimal speed of the mobile robot, the workspace potential field must be established, and environmental information should be obtained to constrain the robot speed. Second, a costmap needs to be established to detect dynamic obstacles, and obstacle avoidance strategies based on the relative motion relationship between dynamic obstacles and the robot should be proposed to realize dynamic obstacle avoidance. Finally, by combining multiple constraints, the collision-free trajectory planning from the start point to the target point is completed, and the mobile robot realizes collision-free smooth motion. Experimental results show that this approach has satisfactory obstacle avoidance planning effects and superior kinematics characteristics and improves the comfort and safety of the mobile robot.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"156 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122485424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-03DOI: 10.1109/ICARM52023.2021.9536073
Bin Fang, Qingchao Wang, Shixin Zhang, Ziwei Xia, F. Sun, Xiao Lu, Yiyong Yang, Licheng Wu
The human hand can adjust stiffness freely to realize different grasp modes, which is beneficial to adapting to different weights of objects. Therefore, humans have been trying different strategies to simulate hand, such as various rigid hand or soft hand, using different structure designs and materials. In this paper, we propose a novel five-finger soft hand. The layer jamming structure is used to increase stiffness and the vision-based tactile sensor is used to provide perception in the soft hand. Through the grasping experiment, the results show that the soft hand can effectively transmit into different grasping modes and adaptively grasp objects of different shapes. Besides, the tactile data is collected by the sensor and a recognition model is built. Through the test, the accuracy is up to 98.75%. In summary, the grasping ability of the soft hand is satisfied, and the combination of tactile sensor and variable stiffness improves performance further.
{"title":"A Novel Humanoid Soft Hand with Variable Stiffness and Multi-modal Perception *","authors":"Bin Fang, Qingchao Wang, Shixin Zhang, Ziwei Xia, F. Sun, Xiao Lu, Yiyong Yang, Licheng Wu","doi":"10.1109/ICARM52023.2021.9536073","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536073","url":null,"abstract":"The human hand can adjust stiffness freely to realize different grasp modes, which is beneficial to adapting to different weights of objects. Therefore, humans have been trying different strategies to simulate hand, such as various rigid hand or soft hand, using different structure designs and materials. In this paper, we propose a novel five-finger soft hand. The layer jamming structure is used to increase stiffness and the vision-based tactile sensor is used to provide perception in the soft hand. Through the grasping experiment, the results show that the soft hand can effectively transmit into different grasping modes and adaptively grasp objects of different shapes. Besides, the tactile data is collected by the sensor and a recognition model is built. Through the test, the accuracy is up to 98.75%. In summary, the grasping ability of the soft hand is satisfied, and the combination of tactile sensor and variable stiffness improves performance further.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124895662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：