Pub Date : 2014-12-18DOI: 10.1109/ICCAS.2014.6987552
Lihua Deng, J. Fei, Changchun Cai
This paper presents a global fast terminal sliding mode (GFTSM) control based on backstepping design for active power filter(APF). Power system is a typical nonlinear system. Its mathematic model can be reduced order by backstepping method. Power system is decomposed into low-level subsystems until finalizing the design of the control law. The System's stability is ensured by Lyapunov framework. The sliding mode control is added into backstepping control to expand the application scope of control. The sliding mode control doesn't need accurate model of system. Therefore robustness of APF control is improved. Global fast terminal control is added into sliding mode control. The nonlinear term of GFTSM control makes the system converge quickly when far from equilibrium point, and the linear term do when near equilibrium point. It makes the system's state tracking error reach zero in finite time. The control speed of system is enhanced. An experiment based on MATLAB/SIMULINK is designed to compare GFTSM backstepping control with pure backstepping control, and experimental results confirm the validity of proposed APF controller for a three-phase three-wire system.
{"title":"Shunt active power filter based on a novel sliding mode backstepping control for three-phase three-wire system","authors":"Lihua Deng, J. Fei, Changchun Cai","doi":"10.1109/ICCAS.2014.6987552","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987552","url":null,"abstract":"This paper presents a global fast terminal sliding mode (GFTSM) control based on backstepping design for active power filter(APF). Power system is a typical nonlinear system. Its mathematic model can be reduced order by backstepping method. Power system is decomposed into low-level subsystems until finalizing the design of the control law. The System's stability is ensured by Lyapunov framework. The sliding mode control is added into backstepping control to expand the application scope of control. The sliding mode control doesn't need accurate model of system. Therefore robustness of APF control is improved. Global fast terminal control is added into sliding mode control. The nonlinear term of GFTSM control makes the system converge quickly when far from equilibrium point, and the linear term do when near equilibrium point. It makes the system's state tracking error reach zero in finite time. The control speed of system is enhanced. An experiment based on MATLAB/SIMULINK is designed to compare GFTSM backstepping control with pure backstepping control, and experimental results confirm the validity of proposed APF controller for a three-phase three-wire system.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"31 1","pages":"1146-1151"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78072818","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6987915
A. Mahmood, Yoonsoo Kim
This paper deals with leader-following formation control and heading synchronization for a group of quadcopters. Unlike most of the existing works that involve a complex design of nonlinear formation and heading control laws for quadcopters, this work proposes a simple linear law especially for quadcopters in leader-following formation. In fact, the quadcopter dynamics involves state variables being coupled nonlinearly. However, it is shown that the dynamics can be treated as a linear system via a feedback linearization scheme. Consequently, a linear formation and heading control law can be designed for the linear system such that (1) the group of quadcopters achieves a desired formation (desired relative altitude and hovering) through local information exchange; (2) follower quadcopters synchronize their heading angles with the leader's; and (3) the leader quadcopter moves along a prescribed path to guide the formation. Although the present work considers quadcopter dynamics only, the proposed methodology can also be applied to various attitude synchronization problems arising in space and robotic applications.
{"title":"Leader-following formation and heading control of networked quadcopters","authors":"A. Mahmood, Yoonsoo Kim","doi":"10.1109/ICCAS.2014.6987915","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987915","url":null,"abstract":"This paper deals with leader-following formation control and heading synchronization for a group of quadcopters. Unlike most of the existing works that involve a complex design of nonlinear formation and heading control laws for quadcopters, this work proposes a simple linear law especially for quadcopters in leader-following formation. In fact, the quadcopter dynamics involves state variables being coupled nonlinearly. However, it is shown that the dynamics can be treated as a linear system via a feedback linearization scheme. Consequently, a linear formation and heading control law can be designed for the linear system such that (1) the group of quadcopters achieves a desired formation (desired relative altitude and hovering) through local information exchange; (2) follower quadcopters synchronize their heading angles with the leader's; and (3) the leader quadcopter moves along a prescribed path to guide the formation. Although the present work considers quadcopter dynamics only, the proposed methodology can also be applied to various attitude synchronization problems arising in space and robotic applications.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"5 1","pages":"919-921"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87699625","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6987940
Yoonseon Oh, Songhwai Oh
Visual information extracted from RGB images has been successfully used for mobile robot localization. The main difficulty with localization using RGB images is that visual features from RGB images are not completely invariant against changes in viewpoints and lighting conditions. This problem can be overcome using features from RGB-D images. In this paper, we evaluate two depth features, depth patches and histograms of oriented normal vectors, extracted from RGB-D images for localization of a mobile robot and demonstrate that robust localization is possible under varying lighting conditions.
{"title":"Robust localization using RGB-D images","authors":"Yoonseon Oh, Songhwai Oh","doi":"10.1109/ICCAS.2014.6987940","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987940","url":null,"abstract":"Visual information extracted from RGB images has been successfully used for mobile robot localization. The main difficulty with localization using RGB images is that visual features from RGB images are not completely invariant against changes in viewpoints and lighting conditions. This problem can be overcome using features from RGB-D images. In this paper, we evaluate two depth features, depth patches and histograms of oriented normal vectors, extracted from RGB-D images for localization of a mobile robot and demonstrate that robust localization is possible under varying lighting conditions.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"21 1","pages":"1023-1026"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86869501","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6987866
C. Nitschke, Y. Minami, Masayuki Hiromoto, Hiroaki Ohshima, Takashi Sato
Unmanned aerial vehicles (UAVs) have many applications and quickly gain popularity with the availability of low-cost micro aerial vehicles (MAVs). Robotics is a popular interdisciplinary education target as it involves understanding and collaboration of several disciplines. Thus, UAVs can serve as an ideal study platform. However, as robotics requires technical background, skills and initial efforts, it is commonly applied in long-term courses. In this paper we successfully exploit the opposite case of robotics in short-term education for students without background, in form of a one-day contest on automatic visual UAV navigation. We provide an extensive survey, and show that existing material and tools do not fit the task and lack in technical aspects. We introduce a novel open-source programming library that comprises programs to guide learning by experience and allow rapid development. It makes contributions to marker-based tracking, with a novel nested-marker design and accurate calibration parameters estimated from 14 Parrot AR.Drone 2.0 front cameras. We show a detailed discussion of the contest results, which represents an extensive user study regarding robotics in education and the effectiveness of the library. The achievement of a steep learning curve for a complex subject has important implications in interdisciplinary design education, as it allows deep understanding of potentials and limitations to facilitate decision-making, unconventional problem solutions and novel applications.
{"title":"A quadrocopter automatic control contest as an example of interdisciplinary design education","authors":"C. Nitschke, Y. Minami, Masayuki Hiromoto, Hiroaki Ohshima, Takashi Sato","doi":"10.1109/ICCAS.2014.6987866","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987866","url":null,"abstract":"Unmanned aerial vehicles (UAVs) have many applications and quickly gain popularity with the availability of low-cost micro aerial vehicles (MAVs). Robotics is a popular interdisciplinary education target as it involves understanding and collaboration of several disciplines. Thus, UAVs can serve as an ideal study platform. However, as robotics requires technical background, skills and initial efforts, it is commonly applied in long-term courses. In this paper we successfully exploit the opposite case of robotics in short-term education for students without background, in form of a one-day contest on automatic visual UAV navigation. We provide an extensive survey, and show that existing material and tools do not fit the task and lack in technical aspects. We introduce a novel open-source programming library that comprises programs to guide learning by experience and allow rapid development. It makes contributions to marker-based tracking, with a novel nested-marker design and accurate calibration parameters estimated from 14 Parrot AR.Drone 2.0 front cameras. We show a detailed discussion of the contest results, which represents an extensive user study regarding robotics in education and the effectiveness of the library. The achievement of a steep learning curve for a complex subject has important implications in interdisciplinary design education, as it allows deep understanding of potentials and limitations to facilitate decision-making, unconventional problem solutions and novel applications.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"28 1","pages":"678-685"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86891059","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6987922
Seongkyun Jeong, Sanguk Lee, Jaehoon Kim
The GNSS(Global Navigation Satellite System) is used widely and generally these days. The satellite navigation system provides navigation service globally and doesn't need an extra cost to use. Thus this system is portrayed as public service. But this system has a weak point. The satellite signal strength is very weak on the ground so this system is vulnerable against interference signal. When the GNSS interference signal is induced, the receiver cannot operate normally. The interference is presented by jamming or spoofing about GNSS signal. The effect of the interference can be done in various field from car navigation to national infrastructure like mobile communication, power plant, etc. ETRI(Electronics and Telecommunications Research Institute) developed GPS Jamming monitoring system for monitoring interference of GNSS signal. The developed GPS Jamming monitoring system has the functions which are jamming & spoofing detection and jamming location with direction finders. Among the technologies of GPS Jamming monitoring system, this paper describes spoofing detection module. This module monitors GPS signal and detects the spoofing signal with measurement and message of receiver. ETRI performed the functional test of developed spoofing detection module through the spoofing signal generation equipment. In this test, we confirm that the spoofing signal is induced in GPS receiver and verify the performance of spoofing detection.
{"title":"Spoofing detection module test of GPS Jamming monitoring system","authors":"Seongkyun Jeong, Sanguk Lee, Jaehoon Kim","doi":"10.1109/ICCAS.2014.6987922","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987922","url":null,"abstract":"The GNSS(Global Navigation Satellite System) is used widely and generally these days. The satellite navigation system provides navigation service globally and doesn't need an extra cost to use. Thus this system is portrayed as public service. But this system has a weak point. The satellite signal strength is very weak on the ground so this system is vulnerable against interference signal. When the GNSS interference signal is induced, the receiver cannot operate normally. The interference is presented by jamming or spoofing about GNSS signal. The effect of the interference can be done in various field from car navigation to national infrastructure like mobile communication, power plant, etc. ETRI(Electronics and Telecommunications Research Institute) developed GPS Jamming monitoring system for monitoring interference of GNSS signal. The developed GPS Jamming monitoring system has the functions which are jamming & spoofing detection and jamming location with direction finders. Among the technologies of GPS Jamming monitoring system, this paper describes spoofing detection module. This module monitors GPS signal and detects the spoofing signal with measurement and message of receiver. ETRI performed the functional test of developed spoofing detection module through the spoofing signal generation equipment. In this test, we confirm that the spoofing signal is induced in GPS receiver and verify the performance of spoofing detection.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"57 1","pages":"1010-1013"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86894092","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6988009
Moon-Youl Park, Jun-seok Yang, Y. Koo, B. Shim, Y. Jeong, E. Kang, Sung-Hyun Han
Recently it is very important to control robot hands more compact and precise sensors in order to increase compensate the grasping capability and to reduce cabling through the finger in the manipulator. As a matter of fact, the miniaturization and cabling harness represents a significant limitation to the design of small sized precise sensor. The main focus of this research is on a robust grasping control of hand fingers, which consists of a flexible multi-fingered with ten joints.
{"title":"A study on gripping control of robotic hand with ten joints for cooperative working","authors":"Moon-Youl Park, Jun-seok Yang, Y. Koo, B. Shim, Y. Jeong, E. Kang, Sung-Hyun Han","doi":"10.1109/ICCAS.2014.6988009","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6988009","url":null,"abstract":"Recently it is very important to control robot hands more compact and precise sensors in order to increase compensate the grasping capability and to reduce cabling through the finger in the manipulator. As a matter of fact, the miniaturization and cabling harness represents a significant limitation to the design of small sized precise sensor. The main focus of this research is on a robust grasping control of hand fingers, which consists of a flexible multi-fingered with ten joints.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"67 1","pages":"307-309"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87067503","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6987942
Chang-Hoi Kim, Hocheol Shin, Kyungmin Jeong
Lizards with a thin and flexible body can easily pass through a narrow path. Their sprawled gait increases their stability and allows them to overcome obstacles and rough terrain. Therefore, lizard bio-mimetic robots are suitable for reconnaissance and detection. The dynamic modeling of a lizard was carried out by weighing each link of a lizard and constructing its 3D shape using micro-CT scanning. We formulated a periodic sprawled gait which is able to describe the gait of lizards. In addition, we generated an upright gait by altering the gait parameters of the sprawled gait. The sprawled and upright gaits of the dynamic lizard model were simulated with various gait parameters using commercial dynamic analyzing software. The stable gait parameters for the stable trotting were obtained through the simulation.
{"title":"Trot gait simulation of four legged robot based on a sprawled gait","authors":"Chang-Hoi Kim, Hocheol Shin, Kyungmin Jeong","doi":"10.1109/ICCAS.2014.6987942","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987942","url":null,"abstract":"Lizards with a thin and flexible body can easily pass through a narrow path. Their sprawled gait increases their stability and allows them to overcome obstacles and rough terrain. Therefore, lizard bio-mimetic robots are suitable for reconnaissance and detection. The dynamic modeling of a lizard was carried out by weighing each link of a lizard and constructing its 3D shape using micro-CT scanning. We formulated a periodic sprawled gait which is able to describe the gait of lizards. In addition, we generated an upright gait by altering the gait parameters of the sprawled gait. The sprawled and upright gaits of the dynamic lizard model were simulated with various gait parameters using commercial dynamic analyzing software. The stable gait parameters for the stable trotting were obtained through the simulation.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"43 1","pages":"1031-1036"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86292287","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6987878
Darsan R. Patel, D. Frank, C. Crane
A control algorithm for allocating control effort to a generic overactuated system is designed. In the case where redundant actuators in the system allow for an infinite number of possible solutions to a desired trajectory, the controller must decide which solution to take. A PD controller is implemented to solve for the desired force and moment of the trajectory. Once the forces and the moments are acquired, a cost function is utilized to help guide the choice of the actuator configuration. This control method is then validated through the simulation of an autonomous surface vehicle that is utilizing an azimuth thruster propulsion system.
{"title":"Controlling an overactuated vehicle with application to an autonomous surface vehicle utilizing azimuth thrusters","authors":"Darsan R. Patel, D. Frank, C. Crane","doi":"10.1109/ICCAS.2014.6987878","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987878","url":null,"abstract":"A control algorithm for allocating control effort to a generic overactuated system is designed. In the case where redundant actuators in the system allow for an infinite number of possible solutions to a desired trajectory, the controller must decide which solution to take. A PD controller is implemented to solve for the desired force and moment of the trajectory. Once the forces and the moments are acquired, a cost function is utilized to help guide the choice of the actuator configuration. This control method is then validated through the simulation of an autonomous surface vehicle that is utilizing an azimuth thruster propulsion system.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"93 1","pages":"745-750"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83857393","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6987834
Rattikorn Sombutkaew, Y. Kumsang, Orachat Chitsobuk
With the massive increment of patients' medical information and images also limitation in transmission bandwidth, it is a challenging task for developing efficient medical information and image encoding techniques for digital picture archiving and communications (PACS). In order to achieve higher encoding efficiency, this research proposes adaptive quantization via fuzzy classified priority mapping. Image statistical characteristics are used as key features for Fuzzy C-mean clustering. The derived priority map is used to identify levels of importance for each image area. The significant candidates of irregular liver tissues, which need special doctor's attention, will be assigned with higher priority than those from the regular ones. The higher the priority, the greater the number of bits assigned for encoding. An analysis of suitable quantization step size has been conducted. With the selection of appropriate quantization parameters for each priority level, the blocking artifacts can be greatly reduced. This results in quality improvement of the reconstructed images while the compression ratio remains reasonably high.
{"title":"Adaptive quantization with Fuzzy C-mean clustering for liver ultrasound compression","authors":"Rattikorn Sombutkaew, Y. Kumsang, Orachat Chitsobuk","doi":"10.1109/ICCAS.2014.6987834","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987834","url":null,"abstract":"With the massive increment of patients' medical information and images also limitation in transmission bandwidth, it is a challenging task for developing efficient medical information and image encoding techniques for digital picture archiving and communications (PACS). In order to achieve higher encoding efficiency, this research proposes adaptive quantization via fuzzy classified priority mapping. Image statistical characteristics are used as key features for Fuzzy C-mean clustering. The derived priority map is used to identify levels of importance for each image area. The significant candidates of irregular liver tissues, which need special doctor's attention, will be assigned with higher priority than those from the regular ones. The higher the priority, the greater the number of bits assigned for encoding. An analysis of suitable quantization step size has been conducted. With the selection of appropriate quantization parameters for each priority level, the blocking artifacts can be greatly reduced. This results in quality improvement of the reconstructed images while the compression ratio remains reasonably high.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"95 1","pages":"521-524"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83951572","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 : 2014-12-18DOI: 10.1109/ICCAS.2014.6987960
Yasser F. O. Mohammad, T. Nishida
Learning from demonstrations (LfD) is gaining more popularity in robotics due to its promise of providing a human-friendly technique for teaching robots new skills by robotics-naive users. The two main approaches to LfD are dynamic motor primitives (DMP) which models demonstrated motions as dynamical systems with the advantage flexibility in changing the motion's starting position, goal or speed and Gaussian Mixture Modelling/ Gaussian Mixture Regression (GMM/GMR) which represents demonstrated motions as mixtures of Gaussians with the advantage of keeping track of the correlations between different dimensions of learned motions and automatic extraction of motion variability along these dimensions. This paper introduces a third approach that relies on symbolization of demonstrated motions by extending the Symbolic Aggregate approXimation (SAX) to handle multiple dimensions of data. The proposed approach is shown through several real-world evaluations to be more resistant to confusing demonstrations that usually arise when action segmentation is automated. The paper also discusses a possible way to combine SAX based LfD withGMM/GMR in order to preserve the advantages of these two approaches while providing superior confusion resistance.
{"title":"Robust learning from demonstrations using multidimensional SAX","authors":"Yasser F. O. Mohammad, T. Nishida","doi":"10.1109/ICCAS.2014.6987960","DOIUrl":"https://doi.org/10.1109/ICCAS.2014.6987960","url":null,"abstract":"Learning from demonstrations (LfD) is gaining more popularity in robotics due to its promise of providing a human-friendly technique for teaching robots new skills by robotics-naive users. The two main approaches to LfD are dynamic motor primitives (DMP) which models demonstrated motions as dynamical systems with the advantage flexibility in changing the motion's starting position, goal or speed and Gaussian Mixture Modelling/ Gaussian Mixture Regression (GMM/GMR) which represents demonstrated motions as mixtures of Gaussians with the advantage of keeping track of the correlations between different dimensions of learned motions and automatic extraction of motion variability along these dimensions. This paper introduces a third approach that relies on symbolization of demonstrated motions by extending the Symbolic Aggregate approXimation (SAX) to handle multiple dimensions of data. The proposed approach is shown through several real-world evaluations to be more resistant to confusing demonstrations that usually arise when action segmentation is automated. The paper also discusses a possible way to combine SAX based LfD withGMM/GMR in order to preserve the advantages of these two approaches while providing superior confusion resistance.","PeriodicalId":6525,"journal":{"name":"2014 14th International Conference on Control, Automation and Systems (ICCAS 2014)","volume":"172 1","pages":"64-71"},"PeriodicalIF":0.0,"publicationDate":"2014-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82938605","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：