Research On Autonomous Obstacle Avoidance For Indoor Mobile Robot

Mobile robot will encounter various obstacles when it moves in unknown environment. How to avoid these obstacles quickly and safely then reach target point smoothly, it is always a hot research spot in the autonomous mobile robot field. This thesis takes AS-R indoor mobile robot as research platform, and do some research on methods for obstacle detection and obstacle avoidance in unknown environment. Main work includes the following aspects:1. In order to solve the problem of mobile robot localization in indoor environment, dead reckoning algorithm is adopted to acquire real-time location and orientation of the robot, and self-localization of the mobile robot is realized in indoor environment.2. Ultrasonic sensors are combined with infrared sensors to detect the distance of obstacles around the mobile robot. For the characteristics of the two types, the adaptive weight fusion estimated algorithm based on multi-sensor data preprocessing is designed to make data fusion, getting the distance of three directions. The accuracy and anti-interference of the system are increased as well.3. The fuzzy logic algorithm is adopted for obstacle avoidance in unknown environment. Because it does not need precise environment model, the Mamdani fuzzy controller for obstacle avoidance is designed in this paper. The distance value of obstacles is fuzzified which is getting from sensors,then it is imported to fuzzy controller. Based on the designed fuzzy rules, the output is direction and angle of robot rotation in obstacle avoidance, which is used to control object.4. The strategy of obstacle avoidance is researched. Fuzzy control theory is combined with the control method based on behavior, and three behaviors are designed including tending to target behavior, autonomous avoidance, walking along the wall. According to the concept of safe-region, the risk level of collision is defined. A competitively collaborative approach based on safe-priority is proposed to coordinate the three behaviors, which guarantees the mobile robot to avoid various obstacles to reach the target point.Finally, the feasibility and validity of the proposed methods are verified through experiments on AS-R indoor mobile robot platform. Experimental results showed: robot can avoid obstacles quickly in unknown environment, which achieves expected effect.