Real-time footstep planning including capture point trajectory optimization for stable biped navigation

Abstract

For stable and efficient biped navigation, a real-time footstep planner taking bipedal dynamics for walking stability control into consideration is proposed. A capture point (CP)-based walking controller is utilized, and footstep planning including reference CP trajectory generation is formulated as an optimization problem. The footstep planning problem is solved using a particle swarm optimization algorithm. The walking period at every footstep is also planned to achieve more effective footstep planning, along with foot placement. Consequently, footstep placement, walking period, and reference CP trajectory for each footstep are optimized by the proposed method. The footstep optimization is performed in real-time without any approximations or precomputations. The effectiveness of the proposed method is demonstrated through experiments in a three-dimensional environment.