Stabilization of biped robot standing on nonstationary plane

Abstract

The problem of stabilization of a biped robot standing on a nonstationary plane is considered in article. Two different algorithms for solving this problem for two different robot models are given. Control goal for both robots was formulated in the same way. Not to fall, the robots have to maintain their center line of mass at center of bearing area. The first algorithm refers to a virtual model with three links and three degrees of freedom. Control of the virtual model is based on a method of inverse dynamics in which the original model is linearized by feedback. For the linearized system PD controller is constructed. The second algorithm refers to a model that is implemented in practice. This bipedal robot was assembled from a robotics kit Bioloid and has six degrees of freedom. To control this robot we use switching regulator. When the angular velocity of the plane on which robot stands is not great, PD controller is running. At high angular velocities of the plane we also add values from gyroscope to this PD controller.