Outdoor Autonomous Mobile Robot Navigation System Research And Design

Autonomous mobile robot has the ability of self-planning, self -organization and adaptivity. Its goal is moving and accomplishing tasks without human assistance or changing environment. So navigation is an important part of mobile robot field. And it is a key technology of mobile robot intelligence. Although our country has made some achievements, we still fall behind some western countries. And we have few robots applied in the real world. So it is of great theoretic and engineering importance to do this research.The main content of mobile navigation system contains mobile robot localization and path-planning. In order to enable mobile robot the abilities of aptitude and autonomy, the mobile robot navigation system is analysed and researched.First of all, the developing history of mobile robot is introduced and the developing status of mobile robot in domestic and overseas is analysed. The main technologies of mobile robot localization and path-planning nowadays is studied . Based on the requirement, a reasonable overall design of outdoor autonomous mobile robot localization and path-planning is presented.Then a mobile robot control program is written to apply the algorithms of navigation conveniently. The design patterns is applied in the program innovatively so the program can cope with changes easily and it is ready for running in another operating system.After that GPS localization and DR navigation principle are analysed and researched. The error of GPS localization is large and there are areas where the GPS data cannot be received. The error of DR is initially small but will be larger and larger as the time goes. So the Kalman filter is used to combine the GPS data and DR data to integrate their advantages. It is shown by the experiment, the filter improves the localization precision greatly.Then the behavior based motion control method is analysed and is applied in the outdoor mobile robot navigation. By using immune network principle, a behavior arbitration system is built to select behavior. The priority of behavior can be dynamically changed depending on the outside circumstances. The drawback of reactive controlling system is considered, and the planning or reasoning module is added to enhance the system's adaptation to environment and flexibility. In order to verify the feasibility of the algorithm, a simulation program is designed in Visual C++ and the path-planning algorithm is evaluated. Experiments demonstrate that in relatively ideal condition the robot can accomplish task using the algorithm. Then the algorithm is tested on the hardware platform and the result is analysed.In the end, a summary is made for the whole research work and the future research plan is presented.