Research On Small Humanoid Robot Gait Planning And Walking Control Based On Virtual Prototype

Humanoid robot is a class of robot which has the shape of human; it can adapt to current human production and living environment, and can also use tools for working like mankind. Since humanoid robot is widely used in many domains such as medical, military and service sector, it has become the research focus of control area in recent years. How to achieve the biped walking stability of humanoid robot is the basis of ensuring its humanoid feature. This paper majors in gait planning and walking control to the virtual prototype of a small humanoid robot, the main work is as follows:The physical structure and geometric parameter have been designed in this paper, and the appropriate freedom has been set for walking. The Pro/E software has been selected for the assemblage of the robot connecting rod modules. A small humanoid robot virtual prototype model has been built in the ADAMS software by adding the corresponding constraints, drivers and friction force.The forward and inverse kinematics model has been built based D-H matrix method, by which the relationship between rod attitude and joint angle can be obtained. The kinematics model of single-leg supporting period and double-leg supporting period has been built for obtaining of the relationship between rod motion and joint driving torque through Lagrange functions.A three-points planning method has been use to interpolated to the key points in robot walking period with the cubic spline interpolation function. The trajectory function of joint angles on time changing has been derived while the robot is walking forward and lateral, and the dynamic walking stability of the planned result has been verified by the ZMP criterion. The planned trajectory curves have been imported to the joint drivers with using of the AKISPL function in ADAMS software. Dynamic simulation experiment shows, this gait planning method is feasible.The PID control system for united simulation has been built through Simulink tool box, and the online control and optimization have been realized with the aim of robot joint angle controlling. Result shows, the robot walking system with the addition of PID controller achieves a better tracking to the changes of joint angles.