Optimal walking trajectory generation for a biped robot usinggenetic algorithm

摘要

Information of suitable additive via-points, which determine theshapes of trajectories, such as velocities and accelerations at thevia-points, are needed for smooth biped walking. Also, incorrectinformation can cause indifferentiability on the trajectory and resultin discontinuous walking motion. In the paper optimal via-points dataare found using a genetic algorithm which minimizes the sum of deviationof velocities and accelerations as well as a jerk. As a result, weobtain the continuity on the whole trajectory interval, the energydistribution through optimum velocities and accelerations. Smooth andstable walking with the leg trajectory optimized by genetic algorithm isshown by walking experiments