Motion Control Of Humanoid Robot

The humanoid robot research began in the late sixties, although it only have 40 years of history, but the research is developing rapidly. The humanoid robot is a high order, nonlinear, multiple degree of freedom, unstable systems.It provides an ideal platform to the research and practice of control theory. The humanoid robot research has attracted many scholars to work on it, and it has become an important direction in the field of robotics.This paper is mainly about the humanoid robot motion control. In the algorithm, we build a robot simulation platform. We also do dynamics, kinematics analysis. The humanoid robot control algorithm works on the simulation platform. In the control system, we use motion controller, CAN card, motor drive and motor to achieve velocity, position, torque control. Concretely, it includes several aspects, such as:1.To the research of the motion control algorithm, we build the robot simulation platform in LabVIEW. First, the robot is divided into pieces, to buid a robot model in the simulation system. Second, set parameters of the robot model refer to the real humanoid robot and import the visual model, add the joints, set up joints movement parameters. Finally, set parameters of the simulation environment and generate the configuration file.2.Build mathematical model of the robot. Forward kinematic is computed with the D-H method. Forward kinematic use the known joint angles to calculate the robot pose. Inverse kinematics is computed with the trigonometric functions. Inverse kinematics model use a known robot pose to calculate the joint angles. And we use the Lagrangian method to do a dynamic research, the dynamics use known robot trajectories to calculate each joint torque.3.Inverted pendulum theory is used in the trajectory for saving energy. COG point theory and ZMP theory are used in dynamics walking stability judgment of the robot. The combination of the inverted pendulum theory and the zero moment point theory, we use the cubic spline interpolation in the robot gait planning to generate the trajectories of joint positions. And joint position trajectories are converted into joint angles trajectories to facilitate motor control. Finally we achieve a stable walking on the simulation platform.4.Use the motion controller, CAN card, motor controller and motor to build a robot control system. We achived speed, position, torque, three motor control methods via the CAN bus communication protocol.