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/ICARM52023.2021.9536077
Jiaxin Lai, Hanyue Lei, Sheng Bao, Liang Du, Jianjun Yuan, Shugen Ma
In this paper, a novel guide robot for blind people based on an elastic rope and a force sensor is proposed, which overcomes the shortcomings of the existing guide robot which uses rigid stick to pull the blind people and brings uncomfortable guiding experience. The robot has the advantages of simple structure, low cost, light weight, foldability and portability. While walking, blind people can adjust the speed of the robot at any time to match his own speed, which can play a reliable and safe following effect. Experiments are carried out in an office and a larger factory. The results show that the feedback force and speed are always kept in a small and stable range during the process of the robot pulling the blind people, so the user experience is good. And blind people can be safely and reliably brought to the target location.
{"title":"Design of a Portable Indoor Guide Robot for Blind People","authors":"Jiaxin Lai, Hanyue Lei, Sheng Bao, Liang Du, Jianjun Yuan, Shugen Ma","doi":"10.1109/ICARM52023.2021.9536077","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536077","url":null,"abstract":"In this paper, a novel guide robot for blind people based on an elastic rope and a force sensor is proposed, which overcomes the shortcomings of the existing guide robot which uses rigid stick to pull the blind people and brings uncomfortable guiding experience. The robot has the advantages of simple structure, low cost, light weight, foldability and portability. While walking, blind people can adjust the speed of the robot at any time to match his own speed, which can play a reliable and safe following effect. Experiments are carried out in an office and a larger factory. The results show that the feedback force and speed are always kept in a small and stable range during the process of the robot pulling the blind people, so the user experience is good. And blind people can be safely and reliably brought to the target location.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"21 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":"117046474","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/ICARM52023.2021.9536166
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/ICARM52023.2021.9536117
Haozhen Chi, Hairong Su, Qinyuan Ren
It has been proved that continued training recovery for the patients who suffer limb movement disorder helps them in therapy protocols. However, the specialty and numbers of therapist often limit the rehabilitation to these patients. Thus, employing robotic equipment to provide the treatment capability is a way out of this dilemma. Considering the compliance and continuous deformation capacity, a robotic exoskeleton actuated by Pneumatic Arti?cial Muscles (PAMs) is designed as the rehabilitation equipment for elbow joint. Moreover, to precisely control the PAM soft actuator of the proposed equipment, a bio-inspired control approach is explored. Finally, several simulations have been conducted to verify the effectiveness of the proposed system.
{"title":"Control of a Soft Medical Exoskeleton via a Bio-inspired Approach","authors":"Haozhen Chi, Hairong Su, Qinyuan Ren","doi":"10.1109/ICARM52023.2021.9536117","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536117","url":null,"abstract":"It has been proved that continued training recovery for the patients who suffer limb movement disorder helps them in therapy protocols. However, the specialty and numbers of therapist often limit the rehabilitation to these patients. Thus, employing robotic equipment to provide the treatment capability is a way out of this dilemma. Considering the compliance and continuous deformation capacity, a robotic exoskeleton actuated by Pneumatic Arti?cial Muscles (PAMs) is designed as the rehabilitation equipment for elbow joint. Moreover, to precisely control the PAM soft actuator of the proposed equipment, a bio-inspired control approach is explored. Finally, several simulations have been conducted to verify the effectiveness of the proposed system.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"25 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":"114922440","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/ICARM52023.2021.9536164
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/ICARM52023.2021.9536207
Xuqiang Qiao, Yinong Li, Ling Zheng, Zhida Zhang, Wei Yang
Predicting vehicle motion states plays a critical role in the planning and control of the automated vehicle. To improve the performance of the predicting, a novel Gaussian process regression (GPR) based on adaptive optimization was proposed. First, a large number of naturalistic driving data from 28 drivers was collected. In order to consider the characteristics of the different kind of drivers, a clustering algorithm method, i.e., the Gaussian mixture model Kullback-Leibler (GMM-KL) divergence was developed. Furthermore, the optimal initialization and the boundary of hyperparameters for each type of driver were obtained on the basis of the classification. To realize the accurate, stable and real-time predicting of the vehicle states, the adaptive optimization Gaussian process regression (AOGPR) was developed, which optimizing and updating the hyperparameters online with the help of the optimal initialization and bounded constraints. Finally, the comparison results between the AOGPR and traditional GPR(TGPR), as well as the Bayesian Network (BN) model were presented. The results showed that the proposed method reached a better accuracy, stability, performance for vehicle states predicting than the TGPR and BN model.
{"title":"Predicting Vehicle States of Naturalistic Driving Data Based on Adaptive Optimization Gaussian Process Regression*","authors":"Xuqiang Qiao, Yinong Li, Ling Zheng, Zhida Zhang, Wei Yang","doi":"10.1109/ICARM52023.2021.9536207","DOIUrl":"https://doi.org/10.1109/ICARM52023.2021.9536207","url":null,"abstract":"Predicting vehicle motion states plays a critical role in the planning and control of the automated vehicle. To improve the performance of the predicting, a novel Gaussian process regression (GPR) based on adaptive optimization was proposed. First, a large number of naturalistic driving data from 28 drivers was collected. In order to consider the characteristics of the different kind of drivers, a clustering algorithm method, i.e., the Gaussian mixture model Kullback-Leibler (GMM-KL) divergence was developed. Furthermore, the optimal initialization and the boundary of hyperparameters for each type of driver were obtained on the basis of the classification. To realize the accurate, stable and real-time predicting of the vehicle states, the adaptive optimization Gaussian process regression (AOGPR) was developed, which optimizing and updating the hyperparameters online with the help of the optimal initialization and bounded constraints. Finally, the comparison results between the AOGPR and traditional GPR(TGPR), as well as the Bayesian Network (BN) model were presented. The results showed that the proposed method reached a better accuracy, stability, performance for vehicle states predicting than the TGPR and BN model.","PeriodicalId":367307,"journal":{"name":"2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)","volume":"99 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":"116800652","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}
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：