Research On Target Grabbing Technology Based On NAO Robot

With the continuous improvement of technology and the aging of the world population,humanoid robots have gradually become an important research direction in the field of robotics,in which target recognition and grasping is one of the most important functions of humanoid robots.Compared with the grasping of industrial robots in a single environment,the grasping environment of humanoid robots is more complex and diverse.In this paper,a vision servo-based target object grasping system is built with the NAO robot as the research platform,and three aspects are studied: arm kinematics,target object recognition and localization,and path planning of the NAO robot.(1)The kinematic model of the NAO robot arm and the linkage coordinate system were established using the D-H parameter method to analyze the forward and inverse kinematics of the NAO robot arm and to establish the kinematic equations.A Monte Carlo algorithm is used to solve the kinematic workspace of the robot arm to lay the theoretical foundation for the subsequent robot grasping.(2)The target object is recognized by using the target recognition method based on color and contour fusion,and the experiments show that the method can avoid the influence of environmental factors in the target recognition process of the NAO robot and can accurately recognize the target object.The monocular ranging and positioning model of NAO robot is established based on the small-aperture imaging principle,and the accuracy of the ranging model is verified through experiments to meet the experimental requirements.(3)An improved RRT* algorithm is used for path planning research.Firstly,the growth of sampling nodes in the RRT* algorithm is changed so that they grow in the direction of the target point,and the adjustable step length strategy and the method of changing the metric function are proposed to solve the problem of the path planning algorithm falling into local minima;secondly,the redundant node elimination detection strategy is used to eliminate the redundant sampling points in the process of random tree expansion;finally,Bessel curves are used to smooth the paths,and the paths are smoothed by Finally,the paths are smoothed by using Bessel curves,and the experiments are conducted by MATLAB simulation platform to verify the feasibility and effectiveness of the improved RRT* algorithm.(4)Combining the theoretical basis of the above three points to construct the NAO robot grasping system,the NAO robot conducted experiments on grasping objects in the environment without and with obstacles respectively,and the experimental results showed that the NAO robot could successfully grasp the target objects.