Background: Due to the refinement of region of the interests (RoIs), two-stage 3D detection algorithms can usually obtain better performance compared with most single-stage detectors. However, most two-stage methods adopt feature connection, to aggregate the grid point features using multi-scale RoI pooling in the second stage. This connection mode does not consider the correlation between multi-scale grid features. Methods: In the first stage, we employ 3D sparse convolution and 2D convolution to fully extract rich semantic features. Then, a small number of coarse RoIs are predicted based region proposal network (RPN) on generated bird’s eye view (BEV) map. After that, we adopt voxel RoI-pooling strategy to aggregate the neighborhood nonempty voxel features of each grid point in RoI in the last two layers of 3D sparse convolution. In this way, we obtain two aggregated features from 3D sparse voxel space for each grid point. Next, we design an attention feature fusion module. This module includes a local and a global attention layer, which can fully integrate the grid point features from different voxel layers. Results: We carry out relevant experiments on the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) dataset. The average precisions of our proposed method are 88.21%, 81.51%, 77.07% on three difficulty levels (easy, moderate, and hard, respectively) for 3D detection, and 92.30%, 90.19%, 86.00% on three difficulty levels (easy, moderate, and hard, respectively) for BEV detection. Conclusions: In this paper, we propose a novel two-stage 3D detection algorithm named Grid Attention Fusion Region-based Convolutional Neural Network (GAF-RCNN) from point cloud. Because we integrate multi-scale RoI grid features with attention mechanism in the refinement stage, different multi-scale features can be better correlated, achieving a competitive level compared with other well tested detection algorithms. This 3D object detection has important implications for robot and cobot technology.
{"title":"GAF-RCNN: Grid attention fusion 3D object detection from point cloud","authors":"Zheng Li, Guofeng Tong, Hao Peng, Mingwei Ma","doi":"10.12688/cobot.17590.1","DOIUrl":"https://doi.org/10.12688/cobot.17590.1","url":null,"abstract":"Background: Due to the refinement of region of the interests (RoIs), two-stage 3D detection algorithms can usually obtain better performance compared with most single-stage detectors. However, most two-stage methods adopt feature connection, to aggregate the grid point features using multi-scale RoI pooling in the second stage. This connection mode does not consider the correlation between multi-scale grid features. Methods: In the first stage, we employ 3D sparse convolution and 2D convolution to fully extract rich semantic features. Then, a small number of coarse RoIs are predicted based region proposal network (RPN) on generated bird’s eye view (BEV) map. After that, we adopt voxel RoI-pooling strategy to aggregate the neighborhood nonempty voxel features of each grid point in RoI in the last two layers of 3D sparse convolution. In this way, we obtain two aggregated features from 3D sparse voxel space for each grid point. Next, we design an attention feature fusion module. This module includes a local and a global attention layer, which can fully integrate the grid point features from different voxel layers. Results: We carry out relevant experiments on the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) dataset. The average precisions of our proposed method are 88.21%, 81.51%, 77.07% on three difficulty levels (easy, moderate, and hard, respectively) for 3D detection, and 92.30%, 90.19%, 86.00% on three difficulty levels (easy, moderate, and hard, respectively) for BEV detection. Conclusions: In this paper, we propose a novel two-stage 3D detection algorithm named Grid Attention Fusion Region-based Convolutional Neural Network (GAF-RCNN) from point cloud. Because we integrate multi-scale RoI grid features with attention mechanism in the refinement stage, different multi-scale features can be better correlated, achieving a competitive level compared with other well tested detection algorithms. This 3D object detection has important implications for robot and cobot technology.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42727294","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}
Ming Li, Ke Yang, J. Qin, J. Zhong, Zipeng Jiang, Qin Su
Background: Navigation and localization are key to the successful execution of autonomous unmanned underwater vehicles (UUVs) in marine environmental monitoring, underwater 3D mapping, and ocean resource surveys. The estimation of the position and the orientation of autonomous UUVs are a long-standing challenging and fundamental problem. As one of the underwater sensors, camera has always been the focus of attention due to its advantages of low cost and rich content information in visibility waters, especially in the fields of visual perception of the underwater environment, target recognition and tracking. At present, the visual real-time pose estimation technology that can be used for UUVs is mainly divided into geometry-based visual positioning algorithms and deep learning-based visual positioning algorithms. Methods: In order to compare the performance of different positioning algorithms and strategies, this paper uses C++ and python, takes the ORB-SLAM3 algorithm and DF-VO algorithm as representatives to conduct a comparative experiment and analysis. Results: The geometry-based algorithm ORB-SLAM3 is less affected by illumination, performs more stably in different underwater environments, and has a shorter calculation time, but its robustness is poor in complex environments. The visual positioning algorithm DF-VO based on deep learning takes longer time to compute, and the positioning accuracy is more easily affected by illumination, especially in dark conditions. However, its robustness is better in unstructured environments such as large-scale image rotation and dynamic object interference. Conclusions: In general, the deep learning-based algorithm is more robust, but multiple deep learning networks make it need more time to compute. The geometry-based method costs less time and is more accurate in low-light and turbid underwater conditions. However, in real underwater situations, these two methods can be connected as binocular vision or methods of multi-sensor combined pose estimation.
{"title":"Comparative study on real-time pose estimation of vision-based unmanned underwater vehicles","authors":"Ming Li, Ke Yang, J. Qin, J. Zhong, Zipeng Jiang, Qin Su","doi":"10.12688/cobot.17642.1","DOIUrl":"https://doi.org/10.12688/cobot.17642.1","url":null,"abstract":"Background: Navigation and localization are key to the successful execution of autonomous unmanned underwater vehicles (UUVs) in marine environmental monitoring, underwater 3D mapping, and ocean resource surveys. The estimation of the position and the orientation of autonomous UUVs are a long-standing challenging and fundamental problem. As one of the underwater sensors, camera has always been the focus of attention due to its advantages of low cost and rich content information in visibility waters, especially in the fields of visual perception of the underwater environment, target recognition and tracking. At present, the visual real-time pose estimation technology that can be used for UUVs is mainly divided into geometry-based visual positioning algorithms and deep learning-based visual positioning algorithms. Methods: In order to compare the performance of different positioning algorithms and strategies, this paper uses C++ and python, takes the ORB-SLAM3 algorithm and DF-VO algorithm as representatives to conduct a comparative experiment and analysis. Results: The geometry-based algorithm ORB-SLAM3 is less affected by illumination, performs more stably in different underwater environments, and has a shorter calculation time, but its robustness is poor in complex environments. The visual positioning algorithm DF-VO based on deep learning takes longer time to compute, and the positioning accuracy is more easily affected by illumination, especially in dark conditions. However, its robustness is better in unstructured environments such as large-scale image rotation and dynamic object interference. Conclusions: In general, the deep learning-based algorithm is more robust, but multiple deep learning networks make it need more time to compute. The geometry-based method costs less time and is more accurate in low-light and turbid underwater conditions. However, in real underwater situations, these two methods can be connected as binocular vision or methods of multi-sensor combined pose estimation.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43063464","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}
Background: As an important part of robot operation, peg-in-hole assembly has problems such as a low degree of automation, a large amount of tasks and low efficiency. It is still a huge challenge for robots to automatically complete assembly tasks because the traditional assembly control policy requires complex analysis of the contact model and it is difficult to build the contact model. The deep reinforcement learning method does not require the establishment of complex contact models, but the long training time and low data utilization efficiency make the training costs very high. Methods: With the aim of addressing the problem of how to accurately obtain the assembly policy and improve the data utilization rate of the robot in the peg-in-hole assembly, we propose the Experience Fusion Proximal Policy Optimization algorithm (EFPPO) based on the Proximal Policy Optimization algorithm (PPO). The algorithm improves the assembly speed and the utilization efficiency of training data by combining force control policy and adding a memory buffer, respectively. Results: We build a single-axis hole assembly system based on the UR5e robotic arm and six-dimensional force sensor in the CoppeliaSim simulation environment to effectively realize the prediction of the assembly environment. Compared with the traditional Deep Deterministic Policy Gradient algorithm (DDPG) and PPO algorithm, the peg-in-hole assembly success rate reaches 100% and the data utilization rate is 125% higher than that of the PPO algorithm. Conclusions: The EFPPO algorithm has a high exploration efficiency. While improving the assembly speed and training speed, the EFPPO algorithm achieves smooth assembly and accurate prediction of the assembly environment.
{"title":"Fast peg-in-hole assembly policy for robots based on experience fusion proximal optimization","authors":"Yu Men, Ligang Jin, Fengming Li, Rui Song","doi":"10.12688/cobot.17579.1","DOIUrl":"https://doi.org/10.12688/cobot.17579.1","url":null,"abstract":"Background: As an important part of robot operation, peg-in-hole assembly has problems such as a low degree of automation, a large amount of tasks and low efficiency. It is still a huge challenge for robots to automatically complete assembly tasks because the traditional assembly control policy requires complex analysis of the contact model and it is difficult to build the contact model. The deep reinforcement learning method does not require the establishment of complex contact models, but the long training time and low data utilization efficiency make the training costs very high. Methods: With the aim of addressing the problem of how to accurately obtain the assembly policy and improve the data utilization rate of the robot in the peg-in-hole assembly, we propose the Experience Fusion Proximal Policy Optimization algorithm (EFPPO) based on the Proximal Policy Optimization algorithm (PPO). The algorithm improves the assembly speed and the utilization efficiency of training data by combining force control policy and adding a memory buffer, respectively. Results: We build a single-axis hole assembly system based on the UR5e robotic arm and six-dimensional force sensor in the CoppeliaSim simulation environment to effectively realize the prediction of the assembly environment. Compared with the traditional Deep Deterministic Policy Gradient algorithm (DDPG) and PPO algorithm, the peg-in-hole assembly success rate reaches 100% and the data utilization rate is 125% higher than that of the PPO algorithm. Conclusions: The EFPPO algorithm has a high exploration efficiency. While improving the assembly speed and training speed, the EFPPO algorithm achieves smooth assembly and accurate prediction of the assembly environment.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46381431","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}
Background: In physical human-robot interaction (pHRI), admittance control is widely used. The most critical thing in admittance control is the configuration of admittance parameters, but a constant admittance value can not meet the needs of interactive indicators smoothness especially. Variable admittance control is a method to overcome this limitation by adjusting the admittance value in real time. This paper proposes a fuzzy Q-learning (FQL) variable admittance control system, which integrates the fuzzy system (FIS) and reinforcement learning method Q-learning. Methods: FIS is used to turn a continuous input state into fuzzy set and Q-learning is used to train the premise strength of fuzzy rules to get the optimal policy of variable admittance value. To verify the performance of this method, an experiment was performed using an AUBO i5 robot. Training trajectory is point-to-point (PTP) trajectory, several interaction variables before and after training by the algorithm are compared to show the validity of algorithm. Results: Experimental results show that the reward converges to a smaller value in about 25 episodes, and the reward of the last five episodes reduces by 68%. The motion trajectory after algorithm training is closer to the ideal min-jerk trajectory and the deviation and mean value of interaction force become smaller. Conclusions: The proposed FQL method can converge in a few episodes and can improve the performance of pHRI by minimizing the jerk based cost function
{"title":"Fuzzy Q-Learning interaction controller design for collaborative robot","authors":"Kaichen Ying, Chen chin-yin, Longxiang Wang","doi":"10.12688/cobot.17595.1","DOIUrl":"https://doi.org/10.12688/cobot.17595.1","url":null,"abstract":"Background: In physical human-robot interaction (pHRI), admittance control is widely used. The most critical thing in admittance control is the configuration of admittance parameters, but a constant admittance value can not meet the needs of interactive indicators smoothness especially. Variable admittance control is a method to overcome this limitation by adjusting the admittance value in real time. This paper proposes a fuzzy Q-learning (FQL) variable admittance control system, which integrates the fuzzy system (FIS) and reinforcement learning method Q-learning. Methods: FIS is used to turn a continuous input state into fuzzy set and Q-learning is used to train the premise strength of fuzzy rules to get the optimal policy of variable admittance value. To verify the performance of this method, an experiment was performed using an AUBO i5 robot. Training trajectory is point-to-point (PTP) trajectory, several interaction variables before and after training by the algorithm are compared to show the validity of algorithm. Results: Experimental results show that the reward converges to a smaller value in about 25 episodes, and the reward of the last five episodes reduces by 68%. The motion trajectory after algorithm training is closer to the ideal min-jerk trajectory and the deviation and mean value of interaction force become smaller. Conclusions: The proposed FQL method can converge in a few episodes and can improve the performance of pHRI by minimizing the jerk based cost function","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42651780","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}
Dianjun Wang, Xiaofan Yang, Ya Chen, Zilong Wang, Zhongkang Song, Zhikun Zhu, Peng Wang
Background: The application of hydrogen-powered vehicles is increasingly widespread, however, the hydrogen-filling process can be dangerous, to ensure both safety and efficiency. A new robotic hydrogen-filling system whose consisting of a hybrid robot combined with an automatic guided vehicle and robotic arm is designed. Methods: An analysis of functional composition of the system was performed, and the hardware scheme was designed. A dual-differential drive AGV and an end effector including a holding jaw and a sucker were designed. According to the system workflow, the control system is divided into four modules. A path planning simulation considering obstacle avoidance is carried out based on improved artificial potential field method and a trajectory planning of the operating arm is completed using source code written in MATLAB. Results: The simulation results show that the automatic guided vehicle can avoid obstacles and move to the specified position. The planed trajectory for robotic arm has certain smoothness, which can be proved that the operating arm can complete the process of grasping the hydrogenation gun. Conclusions: The robotic hydrogen-filling system can replace human beings in most of the work of the hydrogen-filling process, which provides a theoretical basis for automatic hydrogen refueling station.
{"title":"Design of robotic hydrogen-filling system for hydrogen-powered vehicles","authors":"Dianjun Wang, Xiaofan Yang, Ya Chen, Zilong Wang, Zhongkang Song, Zhikun Zhu, Peng Wang","doi":"10.12688/cobot.17597.1","DOIUrl":"https://doi.org/10.12688/cobot.17597.1","url":null,"abstract":"Background: The application of hydrogen-powered vehicles is increasingly widespread, however, the hydrogen-filling process can be dangerous, to ensure both safety and efficiency. A new robotic hydrogen-filling system whose consisting of a hybrid robot combined with an automatic guided vehicle and robotic arm is designed. Methods: An analysis of functional composition of the system was performed, and the hardware scheme was designed. A dual-differential drive AGV and an end effector including a holding jaw and a sucker were designed. According to the system workflow, the control system is divided into four modules. A path planning simulation considering obstacle avoidance is carried out based on improved artificial potential field method and a trajectory planning of the operating arm is completed using source code written in MATLAB. Results: The simulation results show that the automatic guided vehicle can avoid obstacles and move to the specified position. The planed trajectory for robotic arm has certain smoothness, which can be proved that the operating arm can complete the process of grasping the hydrogenation gun. Conclusions: The robotic hydrogen-filling system can replace human beings in most of the work of the hydrogen-filling process, which provides a theoretical basis for automatic hydrogen refueling station.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43457611","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}
Yunfei Xie, Yucong Yang, Donghua Liu, Shuyao Chen, D. Gao, Bihui Tan, Tao Gong, Qiuling Chen, Lei Bi, Tao Liu, Longjiang Deng
Background: Yttrium iron garnet (Y3Fe5O12, YIG) is a prototype magnetic garnet, which possesses the lowest magnetic damping (α) value so far on the earth among all discovered or synthesized materials. This makes it the best candidate for categories of next generation spintronic devices, possessing great application potentials. Methods: A two-step annealing method, with first annealing carried out at a relative low temperature and second annealing at a relatively higher temperature, had been used for the first time to crystallize room temperature sputtered amorphous Y3Fe5O12 (YIG) films on Gd3Ga5O12 (GGG) substrates. The crystalline structure, surface morphology, static and dynamic magnetic properties of the obtained YIG films were characterized through X-ray diffraction (XRD), atomic force microscopy (AFM), vibrating sample magnetometer (VSM) and ferromagnetic resonance (FMR) systems, respectively. Results: It was found that the YIG films obtained via this elaborate annealing method, have a much smoother surface, lower coercivity field, and better dynamic magnetic properties, than that of the YIG films annealed by ordinary one-step approach. Particularly, the ferromagnetic resonance (FMR) linewidth of the best two-step annealed 25 nm YIG film is lower than ~7 Oe at frequency of 10 GHz. Conclusions: Our work clarifies that this two-step annealing approach can effectively improve the quality of the obtained epitaxial YIG films on GGG substrates.
{"title":"Optimizing the quality of epitaxial Y3Fe5O12 thin films via a two-step post-annealing process","authors":"Yunfei Xie, Yucong Yang, Donghua Liu, Shuyao Chen, D. Gao, Bihui Tan, Tao Gong, Qiuling Chen, Lei Bi, Tao Liu, Longjiang Deng","doi":"10.12688/cobot.17583.1","DOIUrl":"https://doi.org/10.12688/cobot.17583.1","url":null,"abstract":"Background: Yttrium iron garnet (Y3Fe5O12, YIG) is a prototype magnetic garnet, which possesses the lowest magnetic damping (α) value so far on the earth among all discovered or synthesized materials. This makes it the best candidate for categories of next generation spintronic devices, possessing great application potentials. Methods: A two-step annealing method, with first annealing carried out at a relative low temperature and second annealing at a relatively higher temperature, had been used for the first time to crystallize room temperature sputtered amorphous Y3Fe5O12 (YIG) films on Gd3Ga5O12 (GGG) substrates. The crystalline structure, surface morphology, static and dynamic magnetic properties of the obtained YIG films were characterized through X-ray diffraction (XRD), atomic force microscopy (AFM), vibrating sample magnetometer (VSM) and ferromagnetic resonance (FMR) systems, respectively. Results: It was found that the YIG films obtained via this elaborate annealing method, have a much smoother surface, lower coercivity field, and better dynamic magnetic properties, than that of the YIG films annealed by ordinary one-step approach. Particularly, the ferromagnetic resonance (FMR) linewidth of the best two-step annealed 25 nm YIG film is lower than ~7 Oe at frequency of 10 GHz. Conclusions: Our work clarifies that this two-step annealing approach can effectively improve the quality of the obtained epitaxial YIG films on GGG substrates.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42868909","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}
Dianjun Wang, Zilong Wang, Ya Chen, Zhiguo Cui, Y. Zhu, Chaofei Wu
Background: In the process of carrying small forgings and other materials, the trajectory error of the 6-degree of freedom light-load collaborative robot will lead to the deviation of forgings placement position. The kinematics analysis and calibration of 6-degree of freedom light load collaborative robot are performed to solve the problem of trajectory error. Methods: The quaternion and cubic spline interpolation methods are adopted to plan the trajectory of the 6-degree of freedom light load collaborative robot. Based on the kinematic error model, the least squares estimation method is adopted to estimate the parameter error of the robot's connecting rod, and the parameter compensation values of each joint are obtained. Results: The kinematic calibration experiment shows that the coordinates of the robot end center are basically consistent with the actual coordinates after compensation, which verifies the rationality of the kinematic model and calibration method. Conclusions: The study lays the theoretical foundation for the trajectory error correction of the light load collaborative robot.
{"title":"Kinematics analysis and calibration of a 6-degree of freedom light load collaborative robot","authors":"Dianjun Wang, Zilong Wang, Ya Chen, Zhiguo Cui, Y. Zhu, Chaofei Wu","doi":"10.12688/cobot.17568.1","DOIUrl":"https://doi.org/10.12688/cobot.17568.1","url":null,"abstract":"Background: In the process of carrying small forgings and other materials, the trajectory error of the 6-degree of freedom light-load collaborative robot will lead to the deviation of forgings placement position. The kinematics analysis and calibration of 6-degree of freedom light load collaborative robot are performed to solve the problem of trajectory error. Methods: The quaternion and cubic spline interpolation methods are adopted to plan the trajectory of the 6-degree of freedom light load collaborative robot. Based on the kinematic error model, the least squares estimation method is adopted to estimate the parameter error of the robot's connecting rod, and the parameter compensation values of each joint are obtained. Results: The kinematic calibration experiment shows that the coordinates of the robot end center are basically consistent with the actual coordinates after compensation, which verifies the rationality of the kinematic model and calibration method. Conclusions: The study lays the theoretical foundation for the trajectory error correction of the light load collaborative robot.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46128934","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}
Background: The five degree of freedom (5-DOF) manipulator greatly improves the machining efficiency and accuracy because of its high flexibility. They see wide application in various automation fields. The dynamic analysis and modeling of manipulators is of great significance to improve the working accuracy of a manipulator. Methods: For the robot task of sheet metal bending, a 5-DOF manipulator based on sliding mode control strategy is designed in this paper. Firstly, the dynamics of the 5-DOF manipulator is analyzed and the dynamic equation is established. Secondly, based on the principle of sliding mode control, a proportional integral (PI) sliding mode control method for 5-DOF manipulator based on nominal model is proposed. Finally, the sliding mode control simulation experiment of 5-DOF manipulator is carried out to verify its stability. Results: The 5-DOF manipulator with PI sliding mode control has a good control effect by overcoming the influence of modeling error due to its strong robustness, and effectively realizes good control stability. Conclusions: The experimental results show that the 5-DOF manipulator has good response speed and stability. The results also suggest that the manipulator can be widely used in complex scenarios such as medical surgery or industrial production line with high safety requirements.
{"title":"Dynamic analysis and sliding mode control method of 5-DOF manipulator","authors":"Q. Jiang, Kai Cai, Ming Ma","doi":"10.12688/cobot.17574.1","DOIUrl":"https://doi.org/10.12688/cobot.17574.1","url":null,"abstract":"Background: The five degree of freedom (5-DOF) manipulator greatly improves the machining efficiency and accuracy because of its high flexibility. They see wide application in various automation fields. The dynamic analysis and modeling of manipulators is of great significance to improve the working accuracy of a manipulator. Methods: For the robot task of sheet metal bending, a 5-DOF manipulator based on sliding mode control strategy is designed in this paper. Firstly, the dynamics of the 5-DOF manipulator is analyzed and the dynamic equation is established. Secondly, based on the principle of sliding mode control, a proportional integral (PI) sliding mode control method for 5-DOF manipulator based on nominal model is proposed. Finally, the sliding mode control simulation experiment of 5-DOF manipulator is carried out to verify its stability. Results: The 5-DOF manipulator with PI sliding mode control has a good control effect by overcoming the influence of modeling error due to its strong robustness, and effectively realizes good control stability. Conclusions: The experimental results show that the 5-DOF manipulator has good response speed and stability. The results also suggest that the manipulator can be widely used in complex scenarios such as medical surgery or industrial production line with high safety requirements.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43128579","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}
X. Zou, Yuting Zhou, Yuhang Zhou, Yukai Xiao, D. Yuan, Gang Xiang
Background: All-terrain vehicles are mostly used in poor driving environments. A key part of the suspension mechanism of all-terrain vehicles, the lower control arm, bears various loads when the vehicle is driving. This component is prone to be fatigue and failure, which affects the performance of the entire vehicle. Therefore, in order to improve the performance of all-terrain vehicles, the fatigue life of the lower control arm was studied based on the measured force load spectrum. Methods: Firstly, the finite element model of the lower control arm is established, the free modal simulation analysis is carried out, and the experimental research is carried out by building a modal test system. Then combining the calculated modal and experimental modal results, the finite element model is verified. Next, through the road load spectrum acquisition test in the automobile proving ground, the force time history of the lower control arm is obtained, and the signal is processed and analyzed to verify the reliability of the force load signal. On this basis, the boundary constraints of the lower control arm are established based on the actual working conditions of the all-terrain vehicle, and the dynamics simulation analysis is carried out with the measured force as input. Finally, according to stress-strain signal in dynamic analysis results, combining the modified local stress-strain method and the Landgrave damage criterion, the fatigue life of the lower control arm is calculated. Results: The minimum fatigue cycle life of the lower control arm on the test roads is 3.56×105 km, and its fatigue life meets the design and use requirements. Conclusions: The result shows that based on the actual driving load spectrum, the actual driving fatigue life can be calculated and forecasted more accurately.
{"title":"Research on fatigue life of all-terrain vehicle control arm based on measured load spectrum","authors":"X. Zou, Yuting Zhou, Yuhang Zhou, Yukai Xiao, D. Yuan, Gang Xiang","doi":"10.12688/cobot.17566.1","DOIUrl":"https://doi.org/10.12688/cobot.17566.1","url":null,"abstract":"Background: All-terrain vehicles are mostly used in poor driving environments. A key part of the suspension mechanism of all-terrain vehicles, the lower control arm, bears various loads when the vehicle is driving. This component is prone to be fatigue and failure, which affects the performance of the entire vehicle. Therefore, in order to improve the performance of all-terrain vehicles, the fatigue life of the lower control arm was studied based on the measured force load spectrum. Methods: Firstly, the finite element model of the lower control arm is established, the free modal simulation analysis is carried out, and the experimental research is carried out by building a modal test system. Then combining the calculated modal and experimental modal results, the finite element model is verified. Next, through the road load spectrum acquisition test in the automobile proving ground, the force time history of the lower control arm is obtained, and the signal is processed and analyzed to verify the reliability of the force load signal. On this basis, the boundary constraints of the lower control arm are established based on the actual working conditions of the all-terrain vehicle, and the dynamics simulation analysis is carried out with the measured force as input. Finally, according to stress-strain signal in dynamic analysis results, combining the modified local stress-strain method and the Landgrave damage criterion, the fatigue life of the lower control arm is calculated. Results: The minimum fatigue cycle life of the lower control arm on the test roads is 3.56×105 km, and its fatigue life meets the design and use requirements. Conclusions: The result shows that based on the actual driving load spectrum, the actual driving fatigue life can be calculated and forecasted more accurately.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45313235","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}
Background: Applying convolution neural networks for large-scale 3D point clouds semantic segmentation is quiet challenging, due to the unordered characteristics of 3D data and the computation burden of large-scale point clouds. Methods: To solve these problems, we designed DPC-Net (Distributed Point Convolution Network). The input point clouds of DPC-Net are partitioned by the K-nearest neighbor strategy and reordered based on Euclidean distance. For reducing computation and memory consumption while retaining critical features, the random sampling strategy is used and a distributed point convolution operation is designed. Our novel convolution method extracts parallel local geometric information including space distance and angle features, respectively. Furthermore, our proposed method could be easily and efficiently embedded into many networks for point clouds semantic segmentation. Results: Extensive experimental results on the Semantic3D and CSPC (Complex Scene Point Cloud) datasets indicate that the proposed DPC-Net not only obtains state-of-the-art performances but also reduces semantic segmentation time. Conclusions: In general, we present an efficient and lightweight deep convolutional network, DPC-Net, which captures local geometric features and local contextual information to predict point labels.
{"title":"DPC-Net: Distributed Point Convolution Network for large-scale point clouds semantic segmentation","authors":"Yu-Ruei Shao, Guofeng Tong, Hao Peng, Mingwei Ma, Jindong Zhang","doi":"10.12688/cobot.17468.1","DOIUrl":"https://doi.org/10.12688/cobot.17468.1","url":null,"abstract":"Background: Applying convolution neural networks for large-scale 3D point clouds semantic segmentation is quiet challenging, due to the unordered characteristics of 3D data and the computation burden of large-scale point clouds. Methods: To solve these problems, we designed DPC-Net (Distributed Point Convolution Network). The input point clouds of DPC-Net are partitioned by the K-nearest neighbor strategy and reordered based on Euclidean distance. For reducing computation and memory consumption while retaining critical features, the random sampling strategy is used and a distributed point convolution operation is designed. Our novel convolution method extracts parallel local geometric information including space distance and angle features, respectively. Furthermore, our proposed method could be easily and efficiently embedded into many networks for point clouds semantic segmentation. Results: Extensive experimental results on the Semantic3D and CSPC (Complex Scene Point Cloud) datasets indicate that the proposed DPC-Net not only obtains state-of-the-art performances but also reduces semantic segmentation time. Conclusions: In general, we present an efficient and lightweight deep convolutional network, DPC-Net, which captures local geometric features and local contextual information to predict point labels.","PeriodicalId":29807,"journal":{"name":"Cobot","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49205558","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：