The Study Of Intelligent Obstacle Avoidance System Based On Behavior For Mobile Robot

With the development of production and science, mobile robot is more and more widely applied in many aspects of life, the environment it's facing also becoming more and more complicated. Early mobile robot research were advocated in known, structured environments and the environment surrounded the mobile robot was accurate understood beforehand. In unstructured environment, lack of accurate and complete knowledge of the environment limited the conventional control method applied to automatic robot. In the complex dynamic changeable environment, the robot must have certain autonomous obstacle-avoiding ability to finish complex task. This study is mainly the design, simulation and experimental at the unknown, unstructured environment of the autonomous mobile robot obstacle avoidance system.Firstly, the obstacle-avoidance problem and development trend of intelligent mobile robots at home and abroad are analyzed in this thesis, then the behavior control methods of mobile robots is summarized, pointed out that the capacity of autonomous obstacle avoidance of mobile robot is not only an important intelligence indicators, but also an important condition for safe driving for intelligent robot. Traditional control algorithm is usually depended on the accurate mathematical model and completed environmental information. When the environment information is scanty or unknown and the mathematical model of the controlled object is hard to determine, the effects of obstacle avoidance is often unsatisfactory.In response to these circumstances, this thesis researched and designed the behavior-based autonomous obstacle-avoiding system. A series of basic behavior is designed, and the fuzzy control method is chosen to control the main behavior. The system established controller for each mode to control the rotation rate and velocity. Through the switching or fusion of each basic behavior, to create new and more sophisticated behavior and to complete the more complicated task. Improved and competed the two common behavior coordinate methods, the behavior competition method and the behavior fusion method. Analyzed the characteristics of the two methods, and pointed out its sphere of application. Several obstacle avoidance simulation experiments is taken in typical obstacles environment to verify the behavior of the improved coordination method has good robustness, real-time, and achieved good results on obstacle avoidance.In order to verify the effectiveness of the algorithms, the obstacle avoidance system simulation software is designed for mobile robot. The ultrasonic sensor is simulated through mathematical means, so that real-time mobile robot can detect changes of the environment. At the same time, the software can also save the real-time data of mobile robot to text file for later data analysis and process optimization. Finally, the behavior-based obstacle avoidance strategy for mobile robot car is transplanted to the main controller of the DSP. The experiments in the single-obstacle and multi-obstacle achieved the desired results, proved the effectiveness of the algorithm further.