Harmonic Function-Based ZMP Trajectory Generation for Nonlinear Motion of Walking Robots

Abstract

The problems of the trajectory generation for the zero moment point (ZMP) and the center of mass of the bipedal walking robot are closely related to a vertical balance of these robots. One of the traditional approaches to solving this problem consists of the sequentially performing footsteps sequence generation, ZMP trajectory generation and the center of mass trajectory generation. Here we propose a method that directly generates ZMP trajectory using a generalized direction curve. This allows simplifying some of the computations associated with the gait generation. In the paper, the proposed algorithm is analyzed for the case when the generalized direction curve is given by a harmonic function. The gaits implementing the motion between two points as well as turning are presented. The tuning of the gait parameters is discussed.