Local Obstacle Avoidance Algorithm And Application Of Mobile Robot Based On Ultrasonic

Mobile robot combines research of multiple fields, such as sensor technology,artificial intelligence, automatic control, and is widely used in industry, agriculture,service and other fields, and is becoming a hot spot of current reasearch. To developautonomous robots and let robot complete tasks in strange environment, obstacleavoidance function is a precondition, and then self-localization is nessary for robotto reach target, so the research of obstacle avoidance algorighm and locationalgorhtim is of great theoretical and practical significance.Because use ultrasonic to detect environment, and experiment on embeddedrobot, so to find an efficient, feasible ultrasonic obstacle avoidance agorithm andself-localization method that is suitable for indoor environment are the mainresearch content of this article.Due to the indoor environment, use dead rckoning positioning method, for themeasurement of heading angle, both electronic compass and gyroscope can measureheading angle, analysis and verify their advantages and disadvantages by robot,compensate the constant drift of gyroscope and establish linear regression equationto get more accurate heading and positioning accuracy. Choose fuzzy control toachieve function of obstacle avoidance, design fuzzy obstacle avoidance controllerand fuzzy rules according to the practical experience, but due to the number offuzzy rules will increase exponentially with the increasing of the input and isdifficult to be implemented on embedded robot platform, so the hierachical fuzzycontroller is designed to ensure the number of rule increases linearly and improvethe real-time performance. For the U-shaped trap problem of the fuzzy obstacleavoidance algorithm, a method based on deviation of heading angle and coordinatesis used to discove and escape trap.Finally, conduct experiments. Conduct robot localization experiment todemonstrate dead reckoning positioning has high precision over short distance, butwith the increase of distance, error also increase. Conduct obstacle avoidance andcompare with the hierarchical fuzzy control algorithm, the result shows hierarchicalfuzzy control algorithm can make full use of environment information and humanexperience without increase in computational complexity. U-shaped trap experimentresult demonstrate the proposed combination of heading angle difference andcoordinate algorithm to discover and escape trap is effective.