Bio-inspired falling motion control for a biped humanoid robot

Abstract

Although consistent stability is desirable, a biped humanoid robot encounters a high risk of falling. Such falls may cause serious damage to both the robot and the environment. This study focuses on this issue and investigates four strategies based on human protective falling motion. These strategies are: "knee flexion", "torso flexion forward", "torso translation backward" and "knee stretched". First, the effectiveness of the strategies for the safe landing is analyzed from an energy variation perspective of the robot system. The four strategies are used to do negative work that reduced the energy of the robot system, thereby reducing the impact velocity of the robot. Then, a simulation study on a human-sized humanoid robot is conducted to assess the influence of the strategies on safe landing. Finally, based on the simulation results for each strategy, a safe falling motion control method is proposed and validated through simulation.