| The humanoid walking robot is a kind of intelligent robot which uses human as a template. Since the humanoid robot can adapt to the human normal living environment, it has many advantages than other kinds of robots. But it has many restrictions, such as no-linear movement, high degree of freedom, coupling joints and small support area. All of these make humanoid robot stable walking control very complicate and difficult. This thesis mainly studied the gait planning method for humanoid robot on the basis of predecessors’work.This thesis used linear inverted pendulum as the research model and proposed the static gait planning method, online gait planning method based on the Capture Point Theory and slope walking gait planning method respectively for flat road walk, external disturbance walk and slope walk. First, the static gait planning determined the step cycle, hip-roll angle range and the step length according to the motion law of linear inverted pendulum. Then the walking process is divided into single support phase and double support phase, and the joints angles are planned on each phase. The Capture Point online gait planning method analyses the robot motion which exposed to the external force from the space orbit energy perspective, then online calculates the capture point location which can help the system back to balance. This online method has been already successful applied to Nao robot. The slope walking gait planning fully considered the road condition influence on the robot, and designed a posture controller to keep the robot upper body upright during the walking process.Creative work of this thesis include the joints angles based on geometric constraints of bipedal robot lower limb are updated in real time which creates conditions for online gait planning; the Capture Point Theory is applied to the robot disturbance walking, and the feasibility is verified through the Nao robot; realized Nao robot down slope stable walking through the upper body posture control. |
PDF Full Text Request