Research On Gait Planning And Disturbance Balance Controll For Nao Humanoid Robot

Humanoid biped robot is a kind of advanced intelligent robots of highly simulation of human legs walking function. Compared to wheeled or tracked robots the humanoid robots have a higher adaptability of the human environment. They have the potential to replace humans in repeated, high strength and high-precision works in hazardous environments. However, the walking system of the biped robot is a inherently unstable system whose dynamic characteristics are very complex with characteristics of multiple variables, strongly coupled, nonlinear and variable structure, so that themselves have instability. In addition, there will be variable disturbances when the biped robot moving in human environments. Studying of gait controlling and disturbance balance on humanoid robots has a great importance in humanoid robots’development. Moreover, the development of biped walking technology will also promote the development of power-type prosthetic, in this paper, Nao humanoid robot platform are applied to study the gait planning and balance control of biped walking robot. Gait planning, dynamic balance control and static balance control of Nao humanoid robot is analysed to provide a reliable basis for further intelligent study of balance control and control and disturbance balance of prosthetic.Firstly, a kind of gait planning of humanoid robot based on the model of inverted pendulum is introduced. Because of the quality of the robot is generally concentrated in the upper body, in this model the robot’s leg is corresponding inverted pendulum’s rod and the torque-driven of the robot’s ankle joints is looked as the basal joint inverted pendulum drive. After the introduction of the principle of linear inverted pendulum model and gait planning procedure the gait of Nao robot is planned. And the gait is tested on the Nao robot platform. Analysis the experiment results demonstrated the validity of this method.Secondly, a control method based on preview control theory is applied in the stability control of the humanoid robot walking. The current posture information of the robot and the future information of reference ZMP are combined with to design the pre-controller containing more future information to control the walking of the robot. By experimental validation on Nao robot platform, the control method can make a humanoid robot better track to achieve the referential ZMP trajectory during walking and walking more stable.Finally, a disturbance balance control of bipedal humanoid robot based on reaction null space principle is introduced. Humanoid robot acceleration and angular velocity are used to calculate the power and direction of the force applied on the robot through the angular momentum theorem. And then the stable model of humanoid robot is established through reaction null space theory. Finally, a hip strategy is applied to deal with the interferential force to make the humanoid robot be able to maintain balance under a certain range of interferential force. And after the disappearance of the interference the robot can restore the original stable posture through hip strategy. Experience on Nao humanoid robot platform demonstrated the validity of the method.