Research And Implementation Of The Middle-sized Robot's Software Architecture And The Self-localization Algorithm

With the background of Robocup, this thesis provides an overview of middle-sized robot classification, development status and research significance at first. Secondly, it also describes the software architecture of competition program and the design and implementation in self-localization method. Finally, we combined with the hardware features of middle-sized robot, provide a method for the middle-sized robot's self-localization based on the white points of mark lines in Robocup environment.Robotics Competition program's develop under the environment of Visual Studio. Net 2003. In order to reduce the complexity of the program, make the program design, debugging and maintenance operations simple, we take the block-based design. In addition, to improve the processing capability and the interaction between modules, this program also uses multi-thread technology.As a hotspot research direction in middle-sized robot field, self-localization method is the centre of competition program. The method for the middle-sized robot's self-localization based on the white points of mark lines is a kind of absolute self-localization technology. To get the more accurate results, this method not only based on the algorithm of program, but also rectified by the code disc.Experiments show that, this self-localization method could get more accurate result. And the error almost no more than 100cm, it well satisfied the requirements of middle-sized robot competition.