Trajectory Planning and Walking Pattern Generation of Humanoid Robot Motion

摘要

Walking trajectory generation for a humanoid robot is a challenging control issue. In this paper, a walking cycle has been recognized considering human motion, and nine simple steps were distinguished in a full step of walking which form motion trajectory, and generates a simplified ZMP motion formulation. This system was used in humanoid robot simulation motion and is achievable easily in walking steps of robot. A minimum DOFs humanoid robot has been considered and geometrical relationships between the robot links were presented by the Denavit-Hartenberg method. The inverse kinematics equations have been solved regarding to extracted ZMP trajectory formula, and constraints in different steps. As a result; angular velocity, acceleration and power of motors were obtained using the relationships and Jacobin. At each step, extracted data were applied on simulated robot in Matlab, and Visual Nastran software. Zero moment point trajectory was evaluated in simulation environment.