Field Robot Navigation, Obstacle Avoidance System Design

Mobile robot is an important branch in robotics, having broad prospects for development, particularly wild environment mobile robot, widely applied to the investigation, detection, military, search and rescue and other special occasions. These applications propose higher requirements for the navigation and obstacle avoidance ability of mobile robot. With DSP5509 embedded development board for control platform, the navigation and obstacle avoidance system was designed for wild environment mobile robot in the thesis.The main research contents are listed as follows:1) According to wild robot functional demands and key technologies, the navigation and obstacle avoidance system overall plan was designed; and a hybrid hierarchical architecture based on behavior was proposed for mobile robot, being convenient for software modular design and upgrading, meanwhile, improving the real-time of control syetem.2) In the thesis, the cause of sensors error were analysed.Electronic compass heading output kalman filter model was established, and simulation result show that improves heading accuracy. The performance of ultrasonic and infrared ranging sensor was tested and analyzed by experimental method; sing least-squares method, measuredt datas was corrected, and improving the accuracy of ranging.3) Multi-ultrasonic sensors signal processing circuit was designed, completing the conversion of sensor output signal from transit time to voltage; meanwhile the behavior-based navigation obstacle avoidance algorithm was studied and the corresponding sensing data collection, navigation and obstacle avoidance, communications software ware designed for the robot syetem.4) In the thesis, the combination of navigation and positioning system, consisting of GPS, electronic compass and photoelectric encoder, was used,by which the navigation algorithm of mobile robot was designed. The obstacle avoidance strategies based on the real-time information of ranging sensors was determined, and based on fuzzy control of the multi-sensor data fusion obstacle avoidance methods was proposed. At last, the simulation proved the method is feasible and effective.