Pub Date : 2004-12-01DOI: 10.1109/RAMECH.2004.1438958
P. V. D. Ven, C. Flanagan, D. Toal, E. Omerdic
In this paper the use of neural networks for the identification of underwater vehicle dynamics is studied. A flexible way of identifying dynamics is desirable for several reasons. The dynamics of underwater craft are highly non-linear and cross coupling between various degrees of freedom normally exists. To date at best empirical models are available to describe these phenomena. On top of this the underwater environment can change drastically as a result of, for example, weather conditions. Due to their ability to adapt for changing circumstances in an online fashion, neural networks offer an interesting alternative for more conventional means of identification. This paper details an identification process using neural networks. To illustrate the performance of this identification process, these neural networks are then used directly or indirectly in a feedforward loop to control the craft in a simulation study.
{"title":"Identification and control of underwater vehicles with the aid of neural networks","authors":"P. V. D. Ven, C. Flanagan, D. Toal, E. Omerdic","doi":"10.1109/RAMECH.2004.1438958","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438958","url":null,"abstract":"In this paper the use of neural networks for the identification of underwater vehicle dynamics is studied. A flexible way of identifying dynamics is desirable for several reasons. The dynamics of underwater craft are highly non-linear and cross coupling between various degrees of freedom normally exists. To date at best empirical models are available to describe these phenomena. On top of this the underwater environment can change drastically as a result of, for example, weather conditions. Due to their ability to adapt for changing circumstances in an online fashion, neural networks offer an interesting alternative for more conventional means of identification. This paper details an identification process using neural networks. To illustrate the performance of this identification process, these neural networks are then used directly or indirectly in a feedforward loop to control the craft in a simulation study.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122378901","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1438038
T. Saleh, Yap Haw Hann, Z. Zhen, A. Mamun, P. Vadakkepat
Conventional design of a mobile robot ensures its stability by keeping the gravity vector through the center of mass inside the structure's polygon of support determined by the contact points between the structure and the ground. This assumption of quasi-static stability fails to hold when the robot moves at high speed as the inertial forces become significant compared to the static gravitational force. On the other hand, the momentum of the moving structure can be exploited to enhance stability if it is dynamically controlled. This principle was exploited to build a gyroscopically stabilized single-wheeled robot by researchers at Carnegie Melon University (CMU). Our design follows the same principle for stability but uses a different mechanism to effect forward and reverse motion.
{"title":"Design of a gyroscopically stabilized single-wheeled robot","authors":"T. Saleh, Yap Haw Hann, Z. Zhen, A. Mamun, P. Vadakkepat","doi":"10.1109/RAMECH.2004.1438038","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438038","url":null,"abstract":"Conventional design of a mobile robot ensures its stability by keeping the gravity vector through the center of mass inside the structure's polygon of support determined by the contact points between the structure and the ground. This assumption of quasi-static stability fails to hold when the robot moves at high speed as the inertial forces become significant compared to the static gravitational force. On the other hand, the momentum of the moving structure can be exploited to enhance stability if it is dynamically controlled. This principle was exploited to build a gyroscopically stabilized single-wheeled robot by researchers at Carnegie Melon University (CMU). Our design follows the same principle for stability but uses a different mechanism to effect forward and reverse motion.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131024348","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1438013
Hiroki Miura, G. Obinata, Atsushi Nakayama, K. Hase, M. Sasaki, T. Iwami, H. Doki
We introduce an optimization method for the trajectory planning of redundant manipulators which achieve a given task with high efficiency, and apply the method to a wheelchair propulsion problem. A GA is used to optimize the redundant variables of a manipulator. Additionally, the procedure of the method does not use any forward dynamics computation. The effectiveness of the proposed algorithm has been shown by applying the method to wheelchair propulsion.
{"title":"Task based approach on trajectory planning with redundant manipulators, and its application to wheelchair propulsion","authors":"Hiroki Miura, G. Obinata, Atsushi Nakayama, K. Hase, M. Sasaki, T. Iwami, H. Doki","doi":"10.1109/RAMECH.2004.1438013","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438013","url":null,"abstract":"We introduce an optimization method for the trajectory planning of redundant manipulators which achieve a given task with high efficiency, and apply the method to a wheelchair propulsion problem. A GA is used to optimize the redundant variables of a manipulator. Additionally, the procedure of the method does not use any forward dynamics computation. The effectiveness of the proposed algorithm has been shown by applying the method to wheelchair propulsion.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"87 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129316676","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1438030
Chungjong Lee, A. Sowmya
Current robotic localization and SLAM algorithms are restricted to simple geometric features such as lines and corners as landmarks. The richness of the information provided by visual perception has not been fully explored. This paper presents a vision based SLAM algorithm which utilizes visual information with minimal prior assumptions.
{"title":"An approach to appearance-based simultaneous localization and map building (SLAM)","authors":"Chungjong Lee, A. Sowmya","doi":"10.1109/RAMECH.2004.1438030","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438030","url":null,"abstract":"Current robotic localization and SLAM algorithms are restricted to simple geometric features such as lines and corners as landmarks. The richness of the information provided by visual perception has not been fully explored. This paper presents a vision based SLAM algorithm which utilizes visual information with minimal prior assumptions.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122129170","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1438988
F. Lin, Po-Hung Shen
An adaptive fuzzy neural network (AFNN) control system is proposed to control the position of the mover of a field-oriented control permanent magnet linear synchronous motor (PMLSM) servo drive system to track periodic reference trajectories in this study. In the proposed AFNN control system, a FNN with accurate approximation capability is employed to approximate the unknown dynamics of the PMLSM, and a robust compensator is proposed to confront the inevitable approximation errors due to finite number of membership functions and disturbances including the friction force. The adaptive learning algorithm that can learn the parameters of the FNN on line is derived using Lyapunov stability theorem. Moreover, to relax the requirement for the value of lumped uncertainty in the robust compensator which comprises a minimum approximation error, optimal parameter vectors, higher-order terms in Taylor series and friction force, an adaptive lumped uncertainty estimation law is investigated. Furthermore, all the control algorithms are implemented in a TMS320C32 DSP-based control computer. The experimental results due to periodic reference trajectories show that the dynamic behaviors of the proposed control systems are robust with regard to uncertainties.
{"title":"A DSP-based permanent magnet linear motor servo drive using adaptive fuzzy-neural-network control","authors":"F. Lin, Po-Hung Shen","doi":"10.1109/RAMECH.2004.1438988","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438988","url":null,"abstract":"An adaptive fuzzy neural network (AFNN) control system is proposed to control the position of the mover of a field-oriented control permanent magnet linear synchronous motor (PMLSM) servo drive system to track periodic reference trajectories in this study. In the proposed AFNN control system, a FNN with accurate approximation capability is employed to approximate the unknown dynamics of the PMLSM, and a robust compensator is proposed to confront the inevitable approximation errors due to finite number of membership functions and disturbances including the friction force. The adaptive learning algorithm that can learn the parameters of the FNN on line is derived using Lyapunov stability theorem. Moreover, to relax the requirement for the value of lumped uncertainty in the robust compensator which comprises a minimum approximation error, optimal parameter vectors, higher-order terms in Taylor series and friction force, an adaptive lumped uncertainty estimation law is investigated. Furthermore, all the control algorithms are implemented in a TMS320C32 DSP-based control computer. The experimental results due to periodic reference trajectories show that the dynamic behaviors of the proposed control systems are robust with regard to uncertainties.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126568677","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1438945
T. Zanma, Y. Takei, M. Ishida
This paper concerns a control problem for a specification which requires a finer description than the resolution of available sensors and actuators. Such a situation may arise when the sensors are kind of limit switches and actuators are on/off control architecture while control specification is to be achieved as accurate as possible. An experimental setup is taken to clarify our problem setup. The experimental setup is a ball position control system in which a small iron ball rotates on a beam. For the experimental setup, we propose a control algorithm based on discrete input and output and continuous state estimation. Finally, the proposed method is applied to the experimental setup to demonstrate its effectiveness as well as numerical simulation.
{"title":"A small iron ball position control by symbolic input and output","authors":"T. Zanma, Y. Takei, M. Ishida","doi":"10.1109/RAMECH.2004.1438945","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438945","url":null,"abstract":"This paper concerns a control problem for a specification which requires a finer description than the resolution of available sensors and actuators. Such a situation may arise when the sensors are kind of limit switches and actuators are on/off control architecture while control specification is to be achieved as accurate as possible. An experimental setup is taken to clarify our problem setup. The experimental setup is a ball position control system in which a small iron ball rotates on a beam. For the experimental setup, we propose a control algorithm based on discrete input and output and continuous state estimation. Finally, the proposed method is applied to the experimental setup to demonstrate its effectiveness as well as numerical simulation.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127230959","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1438950
Qinghua Yang, Libin Zhang, G. Bao, Sheng Xu, J. Ruan
A novel flexible pneumatic actuator FPA is proposed in this paper. The structure and properties of FPA are described. Based on FPA, flexible pneumatic bending joint, flexible pneumatic torsion joint and flexible pneumatic spherical joint are developed. The mathematical models of flexible pneumatic bending joint and flexible pneumatic spherical joint are established. Experiments are carried out to investigate the characteristics of these two joints by utilizing intelligent control techniques. Experimental results demonstrate that flexible pneumatic bending joint and flexible pneumatic spherical joint can meet the designed requirement of application. Finally, the aim and direction of the future work are forecasted.
{"title":"Research on novel flexible pneumatic actuator FPA","authors":"Qinghua Yang, Libin Zhang, G. Bao, Sheng Xu, J. Ruan","doi":"10.1109/RAMECH.2004.1438950","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438950","url":null,"abstract":"A novel flexible pneumatic actuator FPA is proposed in this paper. The structure and properties of FPA are described. Based on FPA, flexible pneumatic bending joint, flexible pneumatic torsion joint and flexible pneumatic spherical joint are developed. The mathematical models of flexible pneumatic bending joint and flexible pneumatic spherical joint are established. Experiments are carried out to investigate the characteristics of these two joints by utilizing intelligent control techniques. Experimental results demonstrate that flexible pneumatic bending joint and flexible pneumatic spherical joint can meet the designed requirement of application. Finally, the aim and direction of the future work are forecasted.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127367930","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1438941
Zhe Tang, Changjiu Zhou, Zeng-qi Sun
Penalty kicking is a specific task for humanoid robots, which is defined by the robotic soccer community. An effective and powerful kicking is a challenging task because of balance. During the period of kicking, the kicking leg moves very quickly and therefore the dynamics should not be ignored. Furthermore, during the kicking, the center of gravity of the robot moves to the front. These two factors greatly influence the balance of the humanoid robot. In this paper, we keep the robot balanced through adjusting the stance leg's ankle joint, both the dynamics and CG changes are considered. The performance and applicability of the proposed method are illustrated both through simulation and real-time control of a humanoid robot developed by the Humanoid Group (www.robo-erectus.org) at Singapore Polytechnic.
点球是人形机器人的一项特殊任务，由机器人足球界定义。由于平衡的关系，有效而有力的踢腿是一项具有挑战性的任务。在踢腿的过程中，踢腿的动作非常快，因此不应该忽视动力学。此外，在踢腿过程中，机器人的重心向前方移动。这两个因素对仿人机器人的平衡有很大的影响。在本文中，我们通过调整站立腿的踝关节来保持机器人的平衡，同时考虑了动力学和CG的变化。通过新加坡理工学院humanoid Group (www.robo-erectus.org)开发的仿人机器人的仿真和实时控制，说明了所提出方法的性能和适用性。
{"title":"Balance of penalty kicking for a biped robot","authors":"Zhe Tang, Changjiu Zhou, Zeng-qi Sun","doi":"10.1109/RAMECH.2004.1438941","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438941","url":null,"abstract":"Penalty kicking is a specific task for humanoid robots, which is defined by the robotic soccer community. An effective and powerful kicking is a challenging task because of balance. During the period of kicking, the kicking leg moves very quickly and therefore the dynamics should not be ignored. Furthermore, during the kicking, the center of gravity of the robot moves to the front. These two factors greatly influence the balance of the humanoid robot. In this paper, we keep the robot balanced through adjusting the stance leg's ankle joint, both the dynamics and CG changes are considered. The performance and applicability of the proposed method are illustrated both through simulation and real-time control of a humanoid robot developed by the Humanoid Group (www.robo-erectus.org) at Singapore Polytechnic.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114034478","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1438951
Yanjie Liu, Lining Sun, Qingxin Meng
This paper presents a new 2-DOF planar parallel robot with high speed and high precision. Harmonic drive is adopted between an AC servomotor and its driven link in each drive joint, and at the same time a rotary encoder feeds the driven link angle into the position controller of each drive joint. Thus, the resolution of the drive joint is improved greatly. However, flexibility of harmonic drive during high-speed motion causes the sharp elastic vibration. Therefore, an acceleration feedback controller is especially designed to increase the damping coefficient and restrain the elastic vibration. Furthermore, the stability of the entire system including the drive joint's position controller and the acceleration feedback controller is proved using a Lyapunov approach. The experiments of trajectory tracking are conducted on the robotic system, and the results validate the proposed control algorithm, contrasted with that obtained by classical PD control without the acceleration feedback loop.
{"title":"Acceleration feedback control of a harmonic drive parallel robot","authors":"Yanjie Liu, Lining Sun, Qingxin Meng","doi":"10.1109/RAMECH.2004.1438951","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1438951","url":null,"abstract":"This paper presents a new 2-DOF planar parallel robot with high speed and high precision. Harmonic drive is adopted between an AC servomotor and its driven link in each drive joint, and at the same time a rotary encoder feeds the driven link angle into the position controller of each drive joint. Thus, the resolution of the drive joint is improved greatly. However, flexibility of harmonic drive during high-speed motion causes the sharp elastic vibration. Therefore, an acceleration feedback controller is especially designed to increase the damping coefficient and restrain the elastic vibration. Furthermore, the stability of the entire system including the drive joint's position controller and the acceleration feedback controller is proved using a Lyapunov approach. The experiments of trajectory tracking are conducted on the robotic system, and the results validate the proposed control algorithm, contrasted with that obtained by classical PD control without the acceleration feedback loop.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114361805","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 : 2004-12-01DOI: 10.1109/RAMECH.2004.1437996
S. Ge, C. Fua, Wei-Ming Liew
This paper presents a method for synthesizing classes of potential functions, which corresponds to commonly used formations, from a general formation potential function by altering the values of a fixed set of parameters. These potential functions may then be used for formation control in swarms of agents. The properties of the potential functions are also examined in this paper. In addition, obstacle avoidance and target tracking behaviors are also implemented on the individual agents. Simulation studies have been carried out to verify the effectiveness of our approach.
{"title":"Swarm formations using the general formation potential function","authors":"S. Ge, C. Fua, Wei-Ming Liew","doi":"10.1109/RAMECH.2004.1437996","DOIUrl":"https://doi.org/10.1109/RAMECH.2004.1437996","url":null,"abstract":"This paper presents a method for synthesizing classes of potential functions, which corresponds to commonly used formations, from a general formation potential function by altering the values of a fixed set of parameters. These potential functions may then be used for formation control in swarms of agents. The properties of the potential functions are also examined in this paper. In addition, obstacle avoidance and target tracking behaviors are also implemented on the individual agents. Simulation studies have been carried out to verify the effectiveness of our approach.","PeriodicalId":252964,"journal":{"name":"IEEE Conference on Robotics, Automation and Mechatronics, 2004.","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114468827","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：