移动机器人自主地图库构建与环境认知
|
摘要
移动机器人的研究涉及到智能控制技术、计算机技术、模式识别以及人工智能等许多学科,其相关技术的研究越来越受到国内外学者的重视。地图构建是移动机器人研究中的核心问题之一,是移动机器人实现自主完成定位、导航等任务的基础。而环境认知是机器人领域的新兴研究热点,也是认知学领域的一项重要研究项目。本文主要研究复合地图库的构建以及基于激光测距仪的环境认知问题。
本文首先对SmartROB-2移动机器人所装配的里程计传感器和激光测距仪传感器的模型进行了讨论,采用Split-Merge算法实现了对原始激光数据点进行直线拟合。通过实际实验结果讨论了这种方法的优点和缺陷并分析了其适用环境。
本文提出了复合地图库的环境描述方法,包括几何地图、拓扑地图和非规则障碍区描述。其中采用基于栅格统计的方法非规则障碍区的原始激光数据点进行预处理,以便去掉室内环境中的动态障碍的影响,之后使用基于网格的共享近邻聚类算法进行点聚类,最后通过对聚类后的点集进行数学分析得到非规则障碍区的环境描述。复合地图库解决了单一地图无法很好地满足机器人自主完成复杂任务的问题,并且可以适应室内环境中几何结构的复杂性和房间内障碍区的非规则性。仿真结果和实际复杂室内环境下的实验结果均验证了这一结论。
针对基于激光数据的环境认知问题,借鉴了认知学和指纹识别的方法,从环境中提取出由多个特征点构成的环境指纹,并提出了同时路径规划与环境匹配(SPAC算法)实现环境指纹的匹配。为减小通路点特征的误差影响,本文采用ICP算法在每次匹配前将环境指纹按照参考模板进行校准。仿真结果验证了所提方法的有效性和实用性。
The research of mobile robot involves a lot of knowledge such as intelligent control, computer science, pattern recognition and artificial intelligence, and the technology in this field has become the focus in the field of robotics and automation. Map building is one of the essential tasks for an autonomous mobile robot's navigation and self-location. Environment cognition is a new research focus in the field of robotics as well as an important research project in cognitive science. This paper researches on composite map base building and environment cognition based on laser range finder.
The paper introduces the model of SmartROB-2 mobile robot's odometer and laser range finder firstly and use Split-Merge algorithm as fitting straight line method. Experiment results in complex indoor environments show the method's effectiveness, practicability and suitable environment.
Composite map base is proposed to describe environment, which includes metric map, topological map and non-regular obstruction area. This paper realizes original laser data preprocessing of non-regular obstruction area using grid-based statistical method and get rid of dynamic obstacles. It also uses the grid-based nearest neighbor clustering algorithm (GNN) to cluster data points and gets the environment description by mathematical analysis on point sets. Composite map base solves that the simple map can not meet the need of robot autonomous complex tasks completing and can adapts the complexity and non-regularity of geometric structure and obstacles in the indoor environment. Simulations and experiment results implemented in complex indoor environments verify the conclusion.
This paper extracts several environment fingerprints composed of feature points by using cognitive science and fingerprint identification for reference. It also proposes the simultaneously path planning and environment cognition algorithm (SPAC) to realize the environment fingerprint matching. ICP algorithm is used to adjust the environment fingerprint to the reference template before each matching. The results of simulation implemented and further data analysis demonstrate the proposed method's validity and practicability.
本文首先对SmartROB-2移动机器人所装配的里程计传感器和激光测距仪传感器的模型进行了讨论,采用Split-Merge算法实现了对原始激光数据点进行直线拟合。通过实际实验结果讨论了这种方法的优点和缺陷并分析了其适用环境。
本文提出了复合地图库的环境描述方法,包括几何地图、拓扑地图和非规则障碍区描述。其中采用基于栅格统计的方法非规则障碍区的原始激光数据点进行预处理,以便去掉室内环境中的动态障碍的影响,之后使用基于网格的共享近邻聚类算法进行点聚类,最后通过对聚类后的点集进行数学分析得到非规则障碍区的环境描述。复合地图库解决了单一地图无法很好地满足机器人自主完成复杂任务的问题,并且可以适应室内环境中几何结构的复杂性和房间内障碍区的非规则性。仿真结果和实际复杂室内环境下的实验结果均验证了这一结论。
针对基于激光数据的环境认知问题,借鉴了认知学和指纹识别的方法,从环境中提取出由多个特征点构成的环境指纹,并提出了同时路径规划与环境匹配(SPAC算法)实现环境指纹的匹配。为减小通路点特征的误差影响,本文采用ICP算法在每次匹配前将环境指纹按照参考模板进行校准。仿真结果验证了所提方法的有效性和实用性。
The research of mobile robot involves a lot of knowledge such as intelligent control, computer science, pattern recognition and artificial intelligence, and the technology in this field has become the focus in the field of robotics and automation. Map building is one of the essential tasks for an autonomous mobile robot's navigation and self-location. Environment cognition is a new research focus in the field of robotics as well as an important research project in cognitive science. This paper researches on composite map base building and environment cognition based on laser range finder.
The paper introduces the model of SmartROB-2 mobile robot's odometer and laser range finder firstly and use Split-Merge algorithm as fitting straight line method. Experiment results in complex indoor environments show the method's effectiveness, practicability and suitable environment.
Composite map base is proposed to describe environment, which includes metric map, topological map and non-regular obstruction area. This paper realizes original laser data preprocessing of non-regular obstruction area using grid-based statistical method and get rid of dynamic obstacles. It also uses the grid-based nearest neighbor clustering algorithm (GNN) to cluster data points and gets the environment description by mathematical analysis on point sets. Composite map base solves that the simple map can not meet the need of robot autonomous complex tasks completing and can adapts the complexity and non-regularity of geometric structure and obstacles in the indoor environment. Simulations and experiment results implemented in complex indoor environments verify the conclusion.
This paper extracts several environment fingerprints composed of feature points by using cognitive science and fingerprint identification for reference. It also proposes the simultaneously path planning and environment cognition algorithm (SPAC) to realize the environment fingerprint matching. ICP algorithm is used to adjust the environment fingerprint to the reference template before each matching. The results of simulation implemented and further data analysis demonstrate the proposed method's validity and practicability.
引文
[1]蔡自兴.机器人学.北京:清华大学出版社,2000.
[2]Thrun S.Robotic mapping:a survey[R].Pittsburgh:School of Computer Science,Carnegie Mllon University,2002.
[3]Biswas R,Limketkai B,Sanner S,et al.Towards object mapping in non-stationary environments with mobile robots[C].Lausanne:International Conference on Intelligent Robots and Systems,2002.
[4]王玉峰.自主移动机器人地图构建、探索及定位研究:(硕士学位论文).大连:大连理工大学,2004.
[5]周凌翔,叶秀清,顾伟康.国内外自主车研究的最新进展.浙江大学信息与电子工程学系,实验室技术报告.
[6]Anhui Cai,Toshio Fukuda,et al.Integration of Distributed Sensing Information in DARS Based on Evidential reasoning.In:Proc.of the 3rd International Symposium on Distributed Autonomous Robotic Systems,1996,268-279.
[7]Daniel Pagac,Eduardo M.Nebot.An Evidential Approach to probabilistic Map-building.In Leo Dorst(ed.) Reasoning with Uncertainty in Robotics Springer-Verlag,1995.164-170
[8]Hans Jacob S,Feder,JohnJ,et al.Adaptive Mobile Robot Navigation and Mapping.The International Journal of Robotics Research.July 1999,11(7):650-668.
[9]Murray Shanahan,Mark Witkowski.Robot Navigation and Map Building with Event Calculus.In Wording Notes of the IJCAI 99 Workshop on Robot Action Planning,1999.1-10.
[10]赵妍,庄严,王伟.基于概率技术的移动机器人地图构建研究现状与进展.中国控制与决策会议,2008.
[11]蔡自兴,邹小兵.移动机器人环境认知理论与技术的研究.机器人,2004,26(1):87-91.
[12]Betke M,Gurvits L.Mobile robot localization using landmarks[J].IEEE Transactions on Robotics & Automation,1997,12(2):251-263.
[13]Mouaddib EI M,Marhic B.Geometrical matching for mobile robot localization[J].IEEE Transactions on Robotics & Automation,2000,16(5):542-552.
[14]Moravec H.Mind Children:The future of Robot and Human Intelligence[M].Cambridge,Massachusetts:Harvard University Press,1988.
[15]Pattie jensfelt,Henrik I.Christensen.Pose Tracking Using Laser Scanning and Minimalistic Environmental models.IEEE Transactions on Robotics and Automation,April 2001,17(2):138-14.
[16]SICK AG Auto Ident.Technical description LMS 2...laser measurement system.Germany:SICK AG Auto Ident,2000.
[17]K.O.Arras,N.Tomatis,et al.Multisensor on-the-fly localization:precisin and reliabity for applicaions.In Robotics and Autonomous System,2001,34(2-3),pp.131-143.
[18]Felix H.Wullschleger,kai O.Arras,Sjur J.Vestli.A flexible exploration framework for map building.Third European Workshop on Advanced Mobile Robots.1999.pp.49 -56.
[19]Ph.BONNIFAIT,G.GARCIA.A Multisensor Localization Algorithm for Mobile Robots and its Real-Time Experimental Validation.IEEE International Conf.on Robotics and Automation.1996.pp.1395-1402.
[20]Nishizawa T,Ohya A,Yuta,S.An implementation of on-board position estimation for a mobile robot-EKF based odometry and laser reflector landmarks detection.IEEE International Conf.on Robotics and Automation.1995.pp.395 - 400.
[21]Borges G.A,Aldon M.J,Gil T.An optimal pose estimator for map-based mobile robot dynamic localization:experimental comparison with the EKF.IEEE International Conf.on Robotics and Automation.2001.pp.1585-1590.
[22]J.Vandorpe,H.Van Brussel,H,Xu.Exact Dynamic Map Building for a Mobile Robot using Geometrical Primitives Produced by a 2D Range Finder.IEEE Int.Conf.on Robotics and Automation,vol.1,1996,901-908..
[23]Samuel T.Pfister,Stergios I.Roumeliotis,Joel W.Burdick.Weighted Line Fitting Algorithms for Mobile Robot Map Building and Efficient Data Representation.IEEE Int.Conf.on Robotics & Automation,2003,121-128
[24]刘敏娟.基于网格的聚类算法分析与研究:(硕士学位论文).郑州:郑州大学,2007.
[25]Ankerstm,Breun Igm,Kr Iegel,KR Iegel H2P,et al.OPTICS:Ordering points to identify the clustering structure[A].In Proc.ACM SIGMOD Int.Conf.on Management of Data (SIGMOD' 99)[C].Philadelphia,PA,1999.49-60.
[26]Macqueen J.Somemethods for classification and analysis of multivariate obsercations [J].Proc 5~(th) Berkeley Symp Math Statist,Prob,1967,1:281-297.
[27]Kaufman L,Rousseeuw PJ.Finding Group in Data:An Introduction to Cluster Analysis [Z].New York:John Wiley & Sons,1990.
[28]Zarrop M B.Optimal experiment design for dynamic system identification[M].Berlin:Springer,1979.
[29]李白男.伪随机信号相关辨识[M].北京:科学出版社,1987.(Li Bainan.PRBS correlacr identification[M].Beijing:Science Press,1987.)
[30]Tomatis N,Nourbakhsh I,Siegwart R.Hybrid simultaneous localization and map building closing the loop with multihypotheses tracking.Proceedings of the IEEE International Conference on Robotics and Automation,2000.Pp.2749-2754.
[31]Tomatis N,Nourbakhsh I,Siegwart R.Hybrid simultaneous localization and map building:a natural integration of topological and metric.Robotics and Autonomous Systems.2003,44(1).Pp.3-14.
[32]Arras K.O,Tomatis N.Improving robustness and precision in mobile robot localization by using laser range finding and monocular vision.In Third European Workshop on Advanced Mobile Robot(Eurobot),Zurich,Switzerland,1999.
[33]Leslie A M,Xu F,Tremoulet P D,etal.Indexing and the object concept:developing "what" and "where" system[J].Trends in Cognitive Sciences,1998,2(1),10-18.
[34]Robert J.Sternberg.认知心理学.杨炳钧,陈燕,邹枝铃译.北京:中国轻工业出版社,2006.
[35]陈倩.基于细节特征的指纹识别算法研究:(硕士学位论文).武汉:武汉理工大学,2008.
[36]N.J.Mitra,N.Gelfand,H.Pottmann and L.Guibas.Registration of Point Cloud Data from a Geometric Optimization Perspective.Eurographics Symposium on Geometry Processing,2004.Pp.23-32.
[37]Zinber,J.Schmidt,H.Niemann.A Refined ICP Algorithm for Robust 3-D Correspondence Estimation.In Proceedings of the IEEE International Conference on Image Processing,Barcelona,Spain,2003.
[38]Faugeras D,Hebert M.The Representation,Recognition,and Location of 3D Objects[J].Int.J.Robotics Res.1986,(5):27-52.
[39]赵晶.指纹识别算法研究:(硕士学位论文).哈尔滨:哈尔滨工程大学,2007.
[40]罗希平,田捷 自动指纹识别中的增强和细节匹配算法[J],软件学报,2002,13(5):946-956
[2]Thrun S.Robotic mapping:a survey[R].Pittsburgh:School of Computer Science,Carnegie Mllon University,2002.
[3]Biswas R,Limketkai B,Sanner S,et al.Towards object mapping in non-stationary environments with mobile robots[C].Lausanne:International Conference on Intelligent Robots and Systems,2002.
[4]王玉峰.自主移动机器人地图构建、探索及定位研究:(硕士学位论文).大连:大连理工大学,2004.
[5]周凌翔,叶秀清,顾伟康.国内外自主车研究的最新进展.浙江大学信息与电子工程学系,实验室技术报告.
[6]Anhui Cai,Toshio Fukuda,et al.Integration of Distributed Sensing Information in DARS Based on Evidential reasoning.In:Proc.of the 3rd International Symposium on Distributed Autonomous Robotic Systems,1996,268-279.
[7]Daniel Pagac,Eduardo M.Nebot.An Evidential Approach to probabilistic Map-building.In Leo Dorst(ed.) Reasoning with Uncertainty in Robotics Springer-Verlag,1995.164-170
[8]Hans Jacob S,Feder,JohnJ,et al.Adaptive Mobile Robot Navigation and Mapping.The International Journal of Robotics Research.July 1999,11(7):650-668.
[9]Murray Shanahan,Mark Witkowski.Robot Navigation and Map Building with Event Calculus.In Wording Notes of the IJCAI 99 Workshop on Robot Action Planning,1999.1-10.
[10]赵妍,庄严,王伟.基于概率技术的移动机器人地图构建研究现状与进展.中国控制与决策会议,2008.
[11]蔡自兴,邹小兵.移动机器人环境认知理论与技术的研究.机器人,2004,26(1):87-91.
[12]Betke M,Gurvits L.Mobile robot localization using landmarks[J].IEEE Transactions on Robotics & Automation,1997,12(2):251-263.
[13]Mouaddib EI M,Marhic B.Geometrical matching for mobile robot localization[J].IEEE Transactions on Robotics & Automation,2000,16(5):542-552.
[14]Moravec H.Mind Children:The future of Robot and Human Intelligence[M].Cambridge,Massachusetts:Harvard University Press,1988.
[15]Pattie jensfelt,Henrik I.Christensen.Pose Tracking Using Laser Scanning and Minimalistic Environmental models.IEEE Transactions on Robotics and Automation,April 2001,17(2):138-14.
[16]SICK AG Auto Ident.Technical description LMS 2...laser measurement system.Germany:SICK AG Auto Ident,2000.
[17]K.O.Arras,N.Tomatis,et al.Multisensor on-the-fly localization:precisin and reliabity for applicaions.In Robotics and Autonomous System,2001,34(2-3),pp.131-143.
[18]Felix H.Wullschleger,kai O.Arras,Sjur J.Vestli.A flexible exploration framework for map building.Third European Workshop on Advanced Mobile Robots.1999.pp.49 -56.
[19]Ph.BONNIFAIT,G.GARCIA.A Multisensor Localization Algorithm for Mobile Robots and its Real-Time Experimental Validation.IEEE International Conf.on Robotics and Automation.1996.pp.1395-1402.
[20]Nishizawa T,Ohya A,Yuta,S.An implementation of on-board position estimation for a mobile robot-EKF based odometry and laser reflector landmarks detection.IEEE International Conf.on Robotics and Automation.1995.pp.395 - 400.
[21]Borges G.A,Aldon M.J,Gil T.An optimal pose estimator for map-based mobile robot dynamic localization:experimental comparison with the EKF.IEEE International Conf.on Robotics and Automation.2001.pp.1585-1590.
[22]J.Vandorpe,H.Van Brussel,H,Xu.Exact Dynamic Map Building for a Mobile Robot using Geometrical Primitives Produced by a 2D Range Finder.IEEE Int.Conf.on Robotics and Automation,vol.1,1996,901-908..
[23]Samuel T.Pfister,Stergios I.Roumeliotis,Joel W.Burdick.Weighted Line Fitting Algorithms for Mobile Robot Map Building and Efficient Data Representation.IEEE Int.Conf.on Robotics & Automation,2003,121-128
[24]刘敏娟.基于网格的聚类算法分析与研究:(硕士学位论文).郑州:郑州大学,2007.
[25]Ankerstm,Breun Igm,Kr Iegel,KR Iegel H2P,et al.OPTICS:Ordering points to identify the clustering structure[A].In Proc.ACM SIGMOD Int.Conf.on Management of Data (SIGMOD' 99)[C].Philadelphia,PA,1999.49-60.
[26]Macqueen J.Somemethods for classification and analysis of multivariate obsercations [J].Proc 5~(th) Berkeley Symp Math Statist,Prob,1967,1:281-297.
[27]Kaufman L,Rousseeuw PJ.Finding Group in Data:An Introduction to Cluster Analysis [Z].New York:John Wiley & Sons,1990.
[28]Zarrop M B.Optimal experiment design for dynamic system identification[M].Berlin:Springer,1979.
[29]李白男.伪随机信号相关辨识[M].北京:科学出版社,1987.(Li Bainan.PRBS correlacr identification[M].Beijing:Science Press,1987.)
[30]Tomatis N,Nourbakhsh I,Siegwart R.Hybrid simultaneous localization and map building closing the loop with multihypotheses tracking.Proceedings of the IEEE International Conference on Robotics and Automation,2000.Pp.2749-2754.
[31]Tomatis N,Nourbakhsh I,Siegwart R.Hybrid simultaneous localization and map building:a natural integration of topological and metric.Robotics and Autonomous Systems.2003,44(1).Pp.3-14.
[32]Arras K.O,Tomatis N.Improving robustness and precision in mobile robot localization by using laser range finding and monocular vision.In Third European Workshop on Advanced Mobile Robot(Eurobot),Zurich,Switzerland,1999.
[33]Leslie A M,Xu F,Tremoulet P D,etal.Indexing and the object concept:developing "what" and "where" system[J].Trends in Cognitive Sciences,1998,2(1),10-18.
[34]Robert J.Sternberg.认知心理学.杨炳钧,陈燕,邹枝铃译.北京:中国轻工业出版社,2006.
[35]陈倩.基于细节特征的指纹识别算法研究:(硕士学位论文).武汉:武汉理工大学,2008.
[36]N.J.Mitra,N.Gelfand,H.Pottmann and L.Guibas.Registration of Point Cloud Data from a Geometric Optimization Perspective.Eurographics Symposium on Geometry Processing,2004.Pp.23-32.
[37]Zinber,J.Schmidt,H.Niemann.A Refined ICP Algorithm for Robust 3-D Correspondence Estimation.In Proceedings of the IEEE International Conference on Image Processing,Barcelona,Spain,2003.
[38]Faugeras D,Hebert M.The Representation,Recognition,and Location of 3D Objects[J].Int.J.Robotics Res.1986,(5):27-52.
[39]赵晶.指纹识别算法研究:(硕士学位论文).哈尔滨:哈尔滨工程大学,2007.
[40]罗希平,田捷 自动指纹识别中的增强和细节匹配算法[J],软件学报,2002,13(5):946-956