基于地形特征的移动机器人定位控制科学与工程专业论文.docxVIP

II II ABSTRACT Robot localization in outdoor environments is a challenging problem because of unstructured terrains. In many outdoor environments, traditional sensors are used to achieve localization. But they are not suitable to use in some conditions. A novel terrain feature localization technique is advanced to allow the robot to identify its location according to the measurable terrain characteristics. First, a terrain inclination map is extracted from a given topographical map along a specific path. Then a cross-section map is obtained along the chosen path. Based on this cross-section map, a terrain inclination or slope map is extracted along that path. A tilt sensor is used to measure the terrain inclination. The encoder is used to measure the velocity of the robot. The extended Kalman filter (EKF) is used to integrate the data from the tilt and the encoder, and to improve the accuracy of the system. The hardware is designed here to demonstrate the system performance. Codes are written for the microcontroller. C language is used to write all of the codes. According to the data packet protocol and transfer protocol, sensor data are transformed into right format and transferred to main microcontroller. The main microcontroller calculates the data following algorithm and outputs the results. Finally, Simulations and the experiments validate the proposed terrain localization technique. Keywords: robot localization, terrain feature, EKF, encoder, DSP III III ACKNOWLEDGEMENTS My current study in Harbin Institute of Technology will soon come to an end. At the completion of my graduation dissertation, I wish to express my sincere appreciation to all those who have offered me invaluable help during my two and a half years of undergraduate life here. Firstly, I would like to express warmest gratitude to my supervisor and advisor, Professor Xiaorui Zhu, not only for his instructive suggestions and valuable comments on the writing of this dissertation, but also for his