Gait Planning And Balance Maintenance Of The Humanoid Robot 'Kong'

The research of Humanoid robots is one of the most active topics in robotic field. In this paper, the development of humanoid robots in the world is introduced and the subject about stable biped gait generating and balance control of the humanoid robot "kong", which is designed by Cyber System and Control Institute of Zhejiang University, is focused on. In the paper, the offline simulation environment for humanoid robot biped gait generating by matlab and Adams is introduced, an on-line generating angular momentum compensation with waist joint is proposed and an optimal on-line walking pattern generation for biped robots is advised.The main work of the paper is listed below:1. The off-line simulation platform with matlab and Adams for the humanoid robot "kong" is structured. The calculation of kinematic/inverse kinematic, angular /linear momentum and ZMP based on linkage models of the humanoid robot of "kong" is fulfilled. The platform can simulate the gait planning and control algorithm of the biped robot, which can make the gait planning work of biped robot more successfully.2. A walking pattern generation method is proposed for biped robots. The key feature of the proposed method is to obtain an optimal walking pattern on-line with the largest stability in the sense of zero moment point (ZMP) subject to the constraints of torque and velocity of the joint actuators. With the aids of a 3-link dynamic model of the robot legs and third-order spline interpolation, the hip and foot trajectories of biped robots are generated and the analytical objective functions of ZMP as well as constraints of joint torques and angular velocities are obtained in terms of a parameter set. A nonlinear optimization procedure is then applied to obtain the optimal parameter set for the generation of walking pattern.3. As the biped robots walk fast, the extra angular moment along perpendicular is brought as the swinging leg. An optimal perpendicular angular momentum compensation algorithm is proposed, which is a complementary method for biped gait generating with ZMP theory. The general idea of this algorithm is using upper-body twisting to compensate the perpendicular angular momentum caused by legged motion of lower-body. This algorithm is proved to be effective by simulation and experiments on the robot "kong".