基于多传感器数据融合的全自主足球机器人全局地图创建
|
摘要
智能机器人正成为继工业机器人后的另一发展热点,作为智能机器人研究的载体,足球机器人的研究现状一定程度上代表并顺应着机器人智能化的发展方向。全自主足球机器人作为足球机器人研究的较高级形态,其对未知环境感知能力以及对在环境感知过程中出现的不确定性的处理能力,代表着机器人在同时定位和地图创建方面技术水平的高低,也是衡量足球机器人智能化程度高低的重要标志。
本论文以RoboCup中型组足球比赛系统为具体研究对象,设计了一个由比赛球队中各机器人自带的所有传感器、机器人上位机、教练机及相关程序组成的多传感器数据融合子系统,使比赛系统具有全局地图构建能力。
全文的主要内容安排如下:
1.介绍了机器人足球比赛研究和数据融合技术研究的意义,并介绍了机器人足球和多传感器数据融合技术在国内外的发展现状及发展趋势;
2.介绍了整个RoboCup中型组比赛系统的体系结构和工作原理,并对整个比赛系统中的视觉子系统、无线通讯子系统、数据融合子系统、决策子系统和运动控制等五个子系统分别作了简要的介绍;
3.详细阐述多传感器数据融合子系统的设计过程,首先分析数据融合子系统的设计要求,然后分别详细讨论了子系统中传感器的选型、融合结构的选择、融合子系统中的数据传输,并在最后就如何实现对子系统中各机器人各传感器的状态进行监控的问题作了研究和介绍。
4.结合多传感器数据融合子系统在RoboCup中型组去求比赛中的应用特点,介绍足球比赛环境的地图表示方法,并对多传感器数据融合子系统在机器人进行自定位、局部地图构建以及比赛系统的全局地图构建时涉及的算法进行了研究和探讨。最后,对算法的在比赛系统中的实际应用进行了实验,实验结果证明了基于数据融合的方法对全局地图创建的有效性。
在论文的最后对全文进行了总结,说明研究的创新点及主要研究成果,同时指出自己的不足和有待进一步研究的问题。
Intelligent robot becomes another research hotspot after the industry robot. As a carrier of the intelligent robots research, football robot conforms the development direction of the intelligentized robot. The autonomous robot soccer tournament system is a senior form in football robot, and The research of sensing unknown environment and processing the dubious situation in autonomous football robot, has been a symbol of the robot's competence in simultaneous localization and mapping, and also been the most important feature of an intelligent mobile robot in its intelligentize degree.
In this thesis, the author took the multi-mobile intelligent robots as research platform, combining the application of middle-size league of RoboCup system, designed a multi-sensors data fusion subsystem which is made up of several robots, all sensors in each robot, coach computer, and correlative program. The data fusion subsystem is used for processing the environment information that collected by all sensors in each robot twice, and finally educe the all information of global map. It can promote the locating precision of system and extend the visible field of robots. This subsystem can meet the need of the performance in real-time and locating precision of competing system.
This thesis including follows content:
1. Introduced the significance of the research in robot soccer competition firstly, and then introduced the history and state of technology in robot soccer tournament system and multi-sensor data fusion system.
2. Introduced the structure and principle of the whole middle-size league of RoboCup tournament system, and then make a simple explain about the five subsystems including vision subsystem, wireless communication subsystem, data fusion subsystem, decision-making subsystem and entity subsystem.
3. Mainly introduced the designing process of data fusion subsystem step by step. Firstly analyzed the designing requirement of data fusion subsystem in middle-size league of RoboCup system, and then explained the choice of sensors that used to obtain environment information and the data fusion structure in subsystem, lastly discussed the management and watch of all states in the soccer tournament system.
4. By analyzing the characteristic of middle-size league RoboCup system, introduced the mapping way of competition environment, and then introduced the correlative fusing arithmetic in self-localization, local map building and global map building. At last, the data fusion subsystem was used to do experiment, and made a analyzing report to the result of experiment.
In the end , a conclusion made to the thesis, made out the points of innovation and the main result of research, at the same time, pointed out the weakness of thesis and the problem that need to improve.
本论文以RoboCup中型组足球比赛系统为具体研究对象,设计了一个由比赛球队中各机器人自带的所有传感器、机器人上位机、教练机及相关程序组成的多传感器数据融合子系统,使比赛系统具有全局地图构建能力。
全文的主要内容安排如下:
1.介绍了机器人足球比赛研究和数据融合技术研究的意义,并介绍了机器人足球和多传感器数据融合技术在国内外的发展现状及发展趋势;
2.介绍了整个RoboCup中型组比赛系统的体系结构和工作原理,并对整个比赛系统中的视觉子系统、无线通讯子系统、数据融合子系统、决策子系统和运动控制等五个子系统分别作了简要的介绍;
3.详细阐述多传感器数据融合子系统的设计过程,首先分析数据融合子系统的设计要求,然后分别详细讨论了子系统中传感器的选型、融合结构的选择、融合子系统中的数据传输,并在最后就如何实现对子系统中各机器人各传感器的状态进行监控的问题作了研究和介绍。
4.结合多传感器数据融合子系统在RoboCup中型组去求比赛中的应用特点,介绍足球比赛环境的地图表示方法,并对多传感器数据融合子系统在机器人进行自定位、局部地图构建以及比赛系统的全局地图构建时涉及的算法进行了研究和探讨。最后,对算法的在比赛系统中的实际应用进行了实验,实验结果证明了基于数据融合的方法对全局地图创建的有效性。
在论文的最后对全文进行了总结,说明研究的创新点及主要研究成果,同时指出自己的不足和有待进一步研究的问题。
Intelligent robot becomes another research hotspot after the industry robot. As a carrier of the intelligent robots research, football robot conforms the development direction of the intelligentized robot. The autonomous robot soccer tournament system is a senior form in football robot, and The research of sensing unknown environment and processing the dubious situation in autonomous football robot, has been a symbol of the robot's competence in simultaneous localization and mapping, and also been the most important feature of an intelligent mobile robot in its intelligentize degree.
In this thesis, the author took the multi-mobile intelligent robots as research platform, combining the application of middle-size league of RoboCup system, designed a multi-sensors data fusion subsystem which is made up of several robots, all sensors in each robot, coach computer, and correlative program. The data fusion subsystem is used for processing the environment information that collected by all sensors in each robot twice, and finally educe the all information of global map. It can promote the locating precision of system and extend the visible field of robots. This subsystem can meet the need of the performance in real-time and locating precision of competing system.
This thesis including follows content:
1. Introduced the significance of the research in robot soccer competition firstly, and then introduced the history and state of technology in robot soccer tournament system and multi-sensor data fusion system.
2. Introduced the structure and principle of the whole middle-size league of RoboCup tournament system, and then make a simple explain about the five subsystems including vision subsystem, wireless communication subsystem, data fusion subsystem, decision-making subsystem and entity subsystem.
3. Mainly introduced the designing process of data fusion subsystem step by step. Firstly analyzed the designing requirement of data fusion subsystem in middle-size league of RoboCup system, and then explained the choice of sensors that used to obtain environment information and the data fusion structure in subsystem, lastly discussed the management and watch of all states in the soccer tournament system.
4. By analyzing the characteristic of middle-size league RoboCup system, introduced the mapping way of competition environment, and then introduced the correlative fusing arithmetic in self-localization, local map building and global map building. At last, the data fusion subsystem was used to do experiment, and made a analyzing report to the result of experiment.
In the end , a conclusion made to the thesis, made out the points of innovation and the main result of research, at the same time, pointed out the weakness of thesis and the problem that need to improve.
引文
[1]张学习.MiroSot足球机器人的研制:[硕士论文].广州:广东工业大学.2003.5
[2]http://www.nljct.zju.edu.cn/robot2003/RoboCup/
[3]王卫华,陈卫东,席裕庚.移动机器人地图创建中的不确定传感器信息处理.自动化学报第29卷第2期.2003年3月
[4]洪炳熔.机器人足球比赛-发展人工智能的里程碑[J].电子世界.2000,4:4-5.
[5]A.K.Mackworth,"On Seeing Robots",in A.Basu and X.Li,editors,Computer vision:systems,Theory,and Applications,pp.1-13,World Scientific Press,Singapore,1993
[6]Middle Size Robot League Rules and Regulations for 2006.MSL Technical Committee 1997-2006
[7]Fenwick J W,Newman P M,Leonard J J.Cooperative Concun'ent Mapping and Localization.IEEE Intenational Conference on Robotics and Automation.Washington,DC,USA,2002,1810-1817.
[8]Rekleitis I,Dukekl G,Milios E.Probabilistic Cooperative Localization and Mapping in Practice.IEEE Intenational Conference on Robotics and Automation.Taipei,Taiwan,2003,1907-1913
[9]王越超,谈大龙.写作机器人学的研究现状与发展.机器人.1998,52(1):69-75
[10]陈海峰.移动机器人基于多传感器数据融合的定位及地图创建研究.大连理工大学硕士论文.2004
[11]王磊.全区域覆盖移动机器人同时定位与地图创建技术的研究.南京理工大学硕士论文.2005
[12]历茂海,洪炳熔,罗荣华.移动机器人的同时定位和地图创建方法.哈尔滨工业大学学报.2004,36(7):874-876
[13]赵宗贵.信心融合技术及其研究动态-研究方向和建议.长沙:数据融合技术研讨会,1995
[14]Waltz E,Llinas J.多传感器数据融合.赵宗贵等译.电子部28所.28-30
[15]Hall D L.Mathematical techniques in multi-sensor data fusion.Norwood:Artech House,1992
[16]Bar-Shalom Y,Fortmann T E.Tracking and Data Fusion.Artech House,1988
[17]Bar-Shalom Y,ed.Multi-target multi-sensor tracking:Advanced Application.Artech House,1990
[18]Bar-Shalom Y,ed.Multi-target multi-sensor tracking:Advanced Application.Artech House,1992
[19]蔡希尧,康耀红,多传感器数据融合的研究现状与发展战略.长沙:数据融合技术研讨会,1995
[20]Subramanyam S,Vadakkepat P,Kim K.Multi-agent centralized control in soccer robots proceeding of 1997 Micro-Robot World Cup Soccer Tournament,1997,6:49-52
[21]Qicheng He,Yimin Yang,Xuexi Zhang,etal.Local mapping for middle-size league of RoboCup.Tianxu Zhang,eds.MIPPR 2007:Automatic Target Recognition and Image Analysis;and Multispectral Image Acquisition.November 2007,wuhan,China.Vol.6786 67862C-1-67862C-1
[22]薛相雷.RoboCup小型组足球机器人决策研究:[硕士论文].长春:长春理工大学,2006
[23]邓旭玥,易建强,赵冬斌.一种全方位移动机器人的运动学分析.机器人.2006,26:49-53
[24]黄永贤,杨宜民.中型组足球机器人运动控制系统的研制.伺服控制.2008.1.46-48
[25]Yanbiao Huang,Yimin Yang,Qicheng He,etal.Research of robot simultaneous localization and mapping in multiple mobile robot system.Tianxu Zhang,eds.MIPPR 2007:Automatic Target Recognition and Image Analysis;and Multispectral Image Acquisition.November 2007,wuhan,China.Vol.678667862B-1-67862B-7
[26]王卫华,陈卫东,席裕庚.基于不确定信息的移动机器人地图创建研究进展[J].机器人,2001,23(6):563-568
[27]李玉榕,郭志疆,蒋静坪等.多传感器融合在移动机器人运动控制中的应用[J].仪器仪表学报,2002.2:106-110
[28]李开生,张慧慧,费仁元等.定位传感器及其融合技术综述[J].计算机自动测量与控制,2001.9:1-3
[29]孙华,陈俊风,吴林.多传感器信心融合技术及其在机器人中的应用.传感器技术.2003年22卷第9期
[30]龚黎明,辜承林.基于PIC18F258系列单片机的嵌入式系统设计.微计算机信心.2004.8:45-46
[31]http://www.prosilica.com/products/ec655.html
[32]http://detail.zol.com.cn/99/98913/param.shtml
[33]刘诗斌,冯晓毅,李宏.基于椭圆假设的电子罗盘误差补偿方法[J].传感器技术,2002,21(10):28-31
[34]http://www.fuanda.com/products/productinfo.asp?id=221
[35]罗志增,蒋静坪.机器人感觉与多信息融合.背景:兵器工业出版社.2003
[36]杨国胜,窦丽华.数据融合及其应用.背景:兵器工业出版社.2004
[37]康耀红.数据融合理论与应用.西安电子科技大学出版社.2006.3.8-13
[38]http://www.macx.cn/a/a.asp?B=100&ID=408612
[39]Dougls E Comer.Intemetworking with TCP/IP,Principles,Protocols,and Architecture.USA:Pretice-Hall International,Inc.Third Edition 1999
[40]Douglas E.Comer.Computer Network And Internets[M].USA:Pretice-Hall International,Inc.1998:273-299
[41]谢希仁.计算机网络(第二版)[M].大连:大连理工大学出版社,1996.
[42]M F Amect,TCP/IP实用技术指南[M].北京:清华大学出版社.1997
[43]Oriolo G,Ulivi G,VendittelliM.Fuzzy maps:A new tool for mobile robot perception and planning[J].Journal of Robotic System,1997,14(3):179-197
[44]Oriolo G,Ulivi G.Realtime map building and navigation for autonomous robots in unknown environments[C].IEEE Trans Systems,Man,and Cybernetics.1998,28(3):316-322
[45]Thrun S,Bucken A.Integrating Grid-based and topologicalmaps for mobile robot navigation[C].In Proceeding Twelfth NCAI,AAAI,1996.
[46]Kuc R.Siegel M W.Physically based simulation model for acoustic sensor robot navigation[C].IEEE Trans Pattern analysis machine Intelligence.1987,9(6):766-778
[47]Elfes A,Moraves H.High Resolution Maps From Wide Angle Sonar[J].IEEE Int ConfRobotics and Automation,1985:116-121
[48]Ohya A,Nagashima Y,Yuta S.Explore unknown environment and map construction using ultrasonic sensing of normal direction of walls[C].IEEE Int ConfRobotics and Automation,1994:485-492
[49]Chong K S,Kleeman Lindsay.Mobile robt map building from an advanced sonar array and accurate odometer[J].International Journal of Robotics Research,1999,18(1):20-36
[50]Kortenkamp D,Weynouth T.Topological mapping for mobile robots using a combination of sonar and vision sensing[C].In Proceeding Twelfth NCAI,AAAI,1994
[51]郑阿奇.聚类算法研究:[硕士论文].南京:南京师范大学.2004
[52]Reid D B.An algorithm for tracking multiple targets.IEEE Trans.Automatic Control.1988,24(6):843-847
[53]石陆魁,何丕廉.一种基于密度的高效聚类算法.计算机应用.2005.8
[54]王翠茹,朵春红.一种改进的基于密度的DBSCAN聚类算法.广西师范大学学报:自然科学版.2007.4
[55]任兴平,何忠龙.改进DBSCAN算法中参数Eps值的确定.现代电子技术.2007.11
[2]http://www.nljct.zju.edu.cn/robot2003/RoboCup/
[3]王卫华,陈卫东,席裕庚.移动机器人地图创建中的不确定传感器信息处理.自动化学报第29卷第2期.2003年3月
[4]洪炳熔.机器人足球比赛-发展人工智能的里程碑[J].电子世界.2000,4:4-5.
[5]A.K.Mackworth,"On Seeing Robots",in A.Basu and X.Li,editors,Computer vision:systems,Theory,and Applications,pp.1-13,World Scientific Press,Singapore,1993
[6]Middle Size Robot League Rules and Regulations for 2006.MSL Technical Committee 1997-2006
[7]Fenwick J W,Newman P M,Leonard J J.Cooperative Concun'ent Mapping and Localization.IEEE Intenational Conference on Robotics and Automation.Washington,DC,USA,2002,1810-1817.
[8]Rekleitis I,Dukekl G,Milios E.Probabilistic Cooperative Localization and Mapping in Practice.IEEE Intenational Conference on Robotics and Automation.Taipei,Taiwan,2003,1907-1913
[9]王越超,谈大龙.写作机器人学的研究现状与发展.机器人.1998,52(1):69-75
[10]陈海峰.移动机器人基于多传感器数据融合的定位及地图创建研究.大连理工大学硕士论文.2004
[11]王磊.全区域覆盖移动机器人同时定位与地图创建技术的研究.南京理工大学硕士论文.2005
[12]历茂海,洪炳熔,罗荣华.移动机器人的同时定位和地图创建方法.哈尔滨工业大学学报.2004,36(7):874-876
[13]赵宗贵.信心融合技术及其研究动态-研究方向和建议.长沙:数据融合技术研讨会,1995
[14]Waltz E,Llinas J.多传感器数据融合.赵宗贵等译.电子部28所.28-30
[15]Hall D L.Mathematical techniques in multi-sensor data fusion.Norwood:Artech House,1992
[16]Bar-Shalom Y,Fortmann T E.Tracking and Data Fusion.Artech House,1988
[17]Bar-Shalom Y,ed.Multi-target multi-sensor tracking:Advanced Application.Artech House,1990
[18]Bar-Shalom Y,ed.Multi-target multi-sensor tracking:Advanced Application.Artech House,1992
[19]蔡希尧,康耀红,多传感器数据融合的研究现状与发展战略.长沙:数据融合技术研讨会,1995
[20]Subramanyam S,Vadakkepat P,Kim K.Multi-agent centralized control in soccer robots proceeding of 1997 Micro-Robot World Cup Soccer Tournament,1997,6:49-52
[21]Qicheng He,Yimin Yang,Xuexi Zhang,etal.Local mapping for middle-size league of RoboCup.Tianxu Zhang,eds.MIPPR 2007:Automatic Target Recognition and Image Analysis;and Multispectral Image Acquisition.November 2007,wuhan,China.Vol.6786 67862C-1-67862C-1
[22]薛相雷.RoboCup小型组足球机器人决策研究:[硕士论文].长春:长春理工大学,2006
[23]邓旭玥,易建强,赵冬斌.一种全方位移动机器人的运动学分析.机器人.2006,26:49-53
[24]黄永贤,杨宜民.中型组足球机器人运动控制系统的研制.伺服控制.2008.1.46-48
[25]Yanbiao Huang,Yimin Yang,Qicheng He,etal.Research of robot simultaneous localization and mapping in multiple mobile robot system.Tianxu Zhang,eds.MIPPR 2007:Automatic Target Recognition and Image Analysis;and Multispectral Image Acquisition.November 2007,wuhan,China.Vol.678667862B-1-67862B-7
[26]王卫华,陈卫东,席裕庚.基于不确定信息的移动机器人地图创建研究进展[J].机器人,2001,23(6):563-568
[27]李玉榕,郭志疆,蒋静坪等.多传感器融合在移动机器人运动控制中的应用[J].仪器仪表学报,2002.2:106-110
[28]李开生,张慧慧,费仁元等.定位传感器及其融合技术综述[J].计算机自动测量与控制,2001.9:1-3
[29]孙华,陈俊风,吴林.多传感器信心融合技术及其在机器人中的应用.传感器技术.2003年22卷第9期
[30]龚黎明,辜承林.基于PIC18F258系列单片机的嵌入式系统设计.微计算机信心.2004.8:45-46
[31]http://www.prosilica.com/products/ec655.html
[32]http://detail.zol.com.cn/99/98913/param.shtml
[33]刘诗斌,冯晓毅,李宏.基于椭圆假设的电子罗盘误差补偿方法[J].传感器技术,2002,21(10):28-31
[34]http://www.fuanda.com/products/productinfo.asp?id=221
[35]罗志增,蒋静坪.机器人感觉与多信息融合.背景:兵器工业出版社.2003
[36]杨国胜,窦丽华.数据融合及其应用.背景:兵器工业出版社.2004
[37]康耀红.数据融合理论与应用.西安电子科技大学出版社.2006.3.8-13
[38]http://www.macx.cn/a/a.asp?B=100&ID=408612
[39]Dougls E Comer.Intemetworking with TCP/IP,Principles,Protocols,and Architecture.USA:Pretice-Hall International,Inc.Third Edition 1999
[40]Douglas E.Comer.Computer Network And Internets[M].USA:Pretice-Hall International,Inc.1998:273-299
[41]谢希仁.计算机网络(第二版)[M].大连:大连理工大学出版社,1996.
[42]M F Amect,TCP/IP实用技术指南[M].北京:清华大学出版社.1997
[43]Oriolo G,Ulivi G,VendittelliM.Fuzzy maps:A new tool for mobile robot perception and planning[J].Journal of Robotic System,1997,14(3):179-197
[44]Oriolo G,Ulivi G.Realtime map building and navigation for autonomous robots in unknown environments[C].IEEE Trans Systems,Man,and Cybernetics.1998,28(3):316-322
[45]Thrun S,Bucken A.Integrating Grid-based and topologicalmaps for mobile robot navigation[C].In Proceeding Twelfth NCAI,AAAI,1996.
[46]Kuc R.Siegel M W.Physically based simulation model for acoustic sensor robot navigation[C].IEEE Trans Pattern analysis machine Intelligence.1987,9(6):766-778
[47]Elfes A,Moraves H.High Resolution Maps From Wide Angle Sonar[J].IEEE Int ConfRobotics and Automation,1985:116-121
[48]Ohya A,Nagashima Y,Yuta S.Explore unknown environment and map construction using ultrasonic sensing of normal direction of walls[C].IEEE Int ConfRobotics and Automation,1994:485-492
[49]Chong K S,Kleeman Lindsay.Mobile robt map building from an advanced sonar array and accurate odometer[J].International Journal of Robotics Research,1999,18(1):20-36
[50]Kortenkamp D,Weynouth T.Topological mapping for mobile robots using a combination of sonar and vision sensing[C].In Proceeding Twelfth NCAI,AAAI,1994
[51]郑阿奇.聚类算法研究:[硕士论文].南京:南京师范大学.2004
[52]Reid D B.An algorithm for tracking multiple targets.IEEE Trans.Automatic Control.1988,24(6):843-847
[53]石陆魁,何丕廉.一种基于密度的高效聚类算法.计算机应用.2005.8
[54]王翠茹,朵春红.一种改进的基于密度的DBSCAN聚类算法.广西师范大学学报:自然科学版.2007.4
[55]任兴平,何忠龙.改进DBSCAN算法中参数Eps值的确定.现代电子技术.2007.11