Research On Automatic Generation Of Robot Arm's Motion Trajectory Of Humanoid 3D Motion

Intelligentization is one of the development goals of robotic technology in the era of artificial intelligence.The advanced goal of robot technology research is to make the robot think and do things like people.If you want the robot to be able to make similar intelligence behavior,it requires the robot to have a strong ability to judge,as well as to learn to gain experience and knowledge in a changing environment.The current technology can not make the robot to meet this requirement,so for specific tasks,users need to program the robot.The programming of the robot not only requires the professional technical knowledge,but also the process is complex and tedious,time-consuming and laborious,which can not meet the expectations given by humans to robots.In order to improve the ability to complete the task and allow the robot to acquire new skills in the process of interacting with people,this topic will study a way to make the robot generate the executable trajectory by imitating the movement of the human arm.Compared with the traditional method of manual programming for the robot,the research of this subject will greatly reduce the difficulty and cycle of programming,improve the automation level of robot programming,enhance the intelligence of human and robot interaction and the ease-using of robot,lay the foundation for the application of robots in a wider field.Firstly,for the problem of how to obtain the data of human arm movement,this paper introduces the principle of multi-camera measurement of 3D spatial point coordinates,and realizes the acquisition of motion data through the learning and use of optical motion capture system.Secondly,aiming at the problem of kinematics modeling of the robot,the kinematics model of the robot is established by DH parameter method with a six-degree-of-freedom manipulator.The DH parameters of the robot and the pose transformation matrix between the links are determined.The positive and inverse kinematics equations are solved.The problem of the multiple solutions of inverse kinematics is discussed,and the solutions are verified by the simulation.Then,for the problem of how to describe the movement of the human body arm,the movement of the arm was analyzed and correspondingly simplified.Using a small amount of joint rotation angle value to describe the arm movement.Based on the fuzzy logic theory,the adaptive neural fuzzy inference system is introduced,and the fuzzy reasoning is carried out by using the neural network learning ability to establish the model of complex nonlinear system.Finally,for the problem of how to generate the robot's trajectory,the obtained arm motion data is converted into the corresponding robot joint path point,and in the joint space,the linear interpolation method with the joint path point and parabolic fitting is used to generate the motion path of the robot.On the basis of the relevant theoretical algorithm,the automatic generation and simulation platform of robot trajectory is prepared.It realizes the real-time acquisition of the movement data of the demonstrators,the correct control of the robot,and the three-dimensional simulation of the generated trajectory.