Dissertation > Excellent graduate degree dissertation topics show

Research on Multiple Robots Cooperative Localization and the Architecture

Author: CaiYunFei
Tutor: TangZhenMin
School: Nanjing University of Technology and Engineering
Course: Pattern Recognition and Intelligent Systems
Keywords: multi-robots cooperative localization mapping service-oriented architecture (SOA) flexible control architecture Extended Kalman Filter (EKF) particle filter (pf) tracked mobile robot
CLC: TP242
Type: PhD thesis
Year: 2011
Downloads: 282
Quote: 0
Read: Download Dissertation

Abstract


In recent years, ground mobile robot technique has been rapidly developed. It has been gradually infiltrating into industry, agriculture, military and some other application domains. As the limitations of the individual robot capabilities, multi-robots collaboration technique is commanded. It can use the advantages of group collaboration to deal with the disadvantages of individual robot and complete the tasks that individual robot can not complete. It is an inevitable trend for robot to imitate human society. It is also an important research topic of artificial intelligence areas and mobile robot technique areas. Cooperative localization is a fundamental issue for multi-robots. The key of study is how to improve their accuracy, efficiency and stability.The method of multi-robots cooperative localization is that robots rely on their own sensors and wireless communication equipments to achieve group localization goals through follow cooperative approaches:such as distributed sensing, information sharing and relative observation. Compared with individual robot localization, multi-robots’cooperation can achieve that sensors detection is wider, information sources are more diverse, and the acuuracy of information fusion is more improved. Addition to the relative position reference, the accuracy of localization and mapping can be greatly improved. It also can be more efficient for cooperation and collbarative control, such as that, the formation can be well kept, the implementation of formation control can be much smoothert. This dissertation is focused on the cooperative system architectures and optimal control methods to improve the cooperation localization accuracy, stability and robustness. Details and results are as follows:1. In order to improve the multi-robots localization accuracy and control stability, Multi-EKF localization and control algorithm is proposed. In the algorithm, the formation model of leader-follower is used. Firstly, each robot is treated as relatively static landmark to the other members. EKF is used to localize itself. Then the formation is considered as a whole unit by the leader robot. The leader robot calculates the joint probability distribution of the whole formation and gets whole formation localization results which are then fed back to individual robot for its localization correction. With the control method of overall forecast-individual correction, the localization accuracy and navigation stability are improved.2. A new hunting strategy based on Cross-EKF localization is proposed in the study of solving the problem of multi-robots cooperative hunting. By this strategy, each hunting robot locates the target with EKF, and the posterior estimate covariance for target estimated by multi-robots is crossly calculated, and a minimum covariance is achieved. The maximum distance from the edge points to the mean center is used as radius to construct a convergence circle. Multi-robots treat the convergence circle as target and round up from all directions. The results show that, if the convergence of dynamic point is extended to the convergence of dynamic surface, it is helpful to improve the convergence rate and control stability of the system.3. A flexible control structure is proposed to solve the problem of the robust control of formation localization and navigation. This structure utilizes the advantages of large redundancy and easy adjustment to enhance the ability of adapting to external dynamic environments. It makes use of the Rao-Blackwellized particle filter to obtain the better association between formation shape and prestress to enhance the carrying ability of the system. And it also uses the role allocation of master-attendant-backup to enhance the system executive performance. The results show that the flexible structure of the multi-robots system can better adapt to the changes of external and internal environments and it has better flexibility and survivability capabilities. The robustness of the system is also improved.4. In the study of online localization and mapping in the process of multi-robots cooperative navigation, multi-robots cooperative online FastSLAM (MRCO-FastSLAM) and the improved MRCO-FastSLAM (IMRCO-FastSLAM) are proposed to solve the problems of accumulative errors in FastSLAM 1.0 for its lack of self-localization-measurements and amendments of algorithm complexity of FastSLAM 2.0 for its self-localization-measurements. On the basis of FastSLAM 1.0 and the leader-follower control model, MRCO-FastSLAM requires the follower robot to provide the ancillary measurement service to the leader robot to help it enhance its own localization estimation accuracy in the process of SLAM. To further improve the accuracy of leader robot localization and reduce the affection of dynamic measurement noise and control noise of follower robot, the improved algorithm named IMRCO-FastSLAM requires the follower robot to give up carrying up SLAM. It uses the method of static observation correction and step-style forward to minimize its self-errors impacts. On the basis of additional waiting time consuming, we use time payment to exchange with space accuracy. In the case of relaxed time requirements, the method has good applicability.5. A new control strategy is proposed to solve the same problem of the robust control of formation localization and navigation based on the study of MRCO-FastSLAM. In this strategy, when the follower robots supply their relative measurements to the leader robot, the posterior estimations of leader robot localization are then as the formation standard location reference for the follower robots to correct their pose prior estimations. On the basis of improving the leader robot’s accuracy, we use the method of relative localization and feedback correction to improve the follower robots’internal localization and control accuracy.6. A layered architecture based on SOA (Service Oriented Architecture) is proposed to construct an efficient, generic and loosely coupled multi-robots architecture for localization and navigation. Service is considered as the constituent elements. SOA packages and interfaces are defined on the normal layered architectures. The lower functional components are encapsulated transparently and the upper service is invoked flexibly during cooperating. The impact of the difference among the heterogeneous robots in cooperation is effectively avoided, which is beneficial to the system construction, expansion, restructuring and maintenance.

Related Dissertations

  1. Query Processing and Optimization in Massive Multi-Database Integration,TP311.13
  2. A Research on Determinating Soil Types and Soil Boundaries under Chinese Soil Taxonomy,S155
  3. Method of mapping UML models to XMI,TP311.5
  4. Genetic Analysis and Molecular Mapping of QTL for Rice Black-Streaked Dwarf Disease Resistance in Rice,S511
  5. Genetic Segregation Analysis and QTL Mapping as Well as Association Analysis for Panicle Angle Trait in Japonica Rice (Oryza Sativa L.),S511.22
  6. New Methodology for Mapping Resistance Trait Loci in Crop Cultivar Population,S336
  7. Mapping of a Wheat Powdery Mildew Resistance Gene in Chinese Landrace Xiaohongpi,S512.1
  8. Association between Leaf Pubescence Status and Resistance to Globular Stink Bug and Mapping Pubescence Gene in Soybean,S565.1
  9. Breeding of Restorer with Erect Pose Panicle and SSR Markers for Alleles of Large Flag Leaf Angle in Japonica Rice (Oryza Sativa L.),S511.22
  10. Genetic Dissection and Elite Allele Identification of Seed Traits in Soybean Cultivars Released from Huanghuai Valleys and Southern China,S565.1
  11. Fine Mapping of Two Leaf Mutant Genes al and yl in Maize,S513
  12. Inheritance of Main Traits Related to Yield and Quality, and Their QTL Mapping in Peanut (Arachis Hypogaea L.),S565.2
  13. Fine Mapping of Inter-Subspecies Hybrid Pollen Sterility Gene in Rice (Oryza Sativa L.) and Phylogenetic Analysis,S511
  14. Genetic Character and Gene Mapping of an Albino Green-Revertible Mutant in Rice (Oryza Sativa L.),S511
  15. Genetic Analysis and Gene Mapping of a Semi-dwarf Mutant in Rice (Oryza Sativa L.),S511
  16. Near-Isogenic Line Development and Precise Mapping of Major Scab Resistance QTL Qfhi.nau-5A in Wheat,S512.1
  17. Nil Development, Precise Mapping and Association Mapping of the Qfhs-3B Scab Resistance QTL in Wheat,S512.1
  18. Fine Mapping of the Blast Resistance Gene Pimh (t) and Detection of Major QTLs in Rice Elite Restorer Line Minghui63,S511.21
  19. QTL Analysis of Salt Tolerance from Solanum Cheesmanii during Seed Germination and Seedling Stages Using Advanced Backcross Population,S641.2
  20. Construction of Molecular Genetic Map and Gene Mapping of Fruit Skin Color for Pyrus Pyrifolia Nakai,S661.2
  21. Identification and Molecular Mapping of New Powdery Mildew Resistance Genes in Common Wheat,S512.1

CLC: > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > Robotics > Robot
© 2012 www.DissertationTopic.Net  Mobile