Design Of Arm-wheeled Teaching-robot And Research On Obstacle Avoidance

To adapt to the rapidly development of robot technology, researches on teaching robot has become one of the most popular themes in institutes and robot companies. This paper is aimed to design a type of wheeled-arm teaching-robot, which has already been tested on teaching experiments and reached an ideal result.The ability to avoid obstacles is one of the most important factors related to usage scale of teaching-robot. In order to improve the obstacle avoidance ability of the teaching-robot, information fusion technology between dynamic neural network and ultrasonic sensor have been introduced to improve the ability of the robot to identify different types of environment in movement. As a result, the complexity of algorithm of obstacle avoidance has been reduced.Then an amendment based on obstacle recognition system and contour extraction algorithm is proposed, which is combined with location algorithm of ultrasonic sensor to make teaching-robot to achieve precise positioning of obstacles.To make up to the disadvantages of teaching robot in unknown environment, a type of fuzzy control based on information fusion is designed to help arm-wheeled robot to make better performance in complex environment though MATLAB&Simulating experiments.The originalities of this paper are listed as follow: The designed robot with a structure of wheeled-arm could achieve different teaching aims during multifunction such as movement, obstacle avoidance and grasp. It has proved to be a befitting teaching appliance in actual courses and has richer functionality compared with simple wheeled-robot or mechanical armed-robot. With the techniques of machine vision information fusion and fuzzy control technology, the designed robot performs better in obstacle avoidance; therefore it is better suitable for teaching.