| The intelligent service of the robot requires the robot to serve in different scenes.This thesis combines the NAO robot platform to develop three single scene modes(visual line-following navigation,object location and grabbing,and moving object tracking and obstacle avoidance),and finally through the frame based expert system to form a knowledge base of each mode,and uses the inference engine to achieve multimode fusion.In the development of visual line-following navigation,the proposed fast path extraction processing method can not only accurately extract the path information in the case of noisy interference,but also improve the running speed to ensure the realtime performance of the robot.A slope compensation PID controller is proposed to control the parameters of the robot during walking,which ensures low error and high stability when the robot follows the navigation line.In the development of object positioning and grabbing,a monocular localization algorithm is proposed using the characteristics of QR code.Then the kinematics of the robot arm is modeled.Finally the end effector of the robot is moved to the specified position according to the three-dimensional coordinates of the QR code to grasp the target object.In the development of moving object tracking and obstacle avoidance,the NaoMark marker is used to simplify the characteristics of the moving target,and the parameters of the robot walking are controlled according to the central position of the marker in the image.In order to avoid the collision of the robot with the obstacle during the movement,the obstacle avoidance is realized by combining the ultrasonic sensor measurement and the artificial potential field algorithm.Finally,the implementation methods of each part of the above three single modes are formed into a knowledge base in the form of a framework based expert system,and the knowledge is combined by the inference engine to realize the fusion of multiple modes to complete tasks that require multiple single mode. |
PDF Full Text Request