Frequency-Temporal Disagreement Adaptation for Robotic Terrain Classification via Vibration in a Dynamic Environment

Gyro-Sensor-Based Vibration Control for Dynamic Humanoid-Robot Walking on Inclined Surfaces

An efficient motor-control system for stable walking of the lightweight humanoid robot KONDO KHR-3HV on inclined surfaces is investigated. The motor-control system is based on the angular velocity of the pitch motion of the robot, which is detected by a gyro sensor attached to the robot torso and referred to as the angular-pitch velocity. The robot gait is analyzed for different downslopes with and without the motor-feedback control. A novel method of frequency-domain analysis of the angular-pitch velocity is proposed for explaining the reasons behind the instabilities of dynamic humanoid-robot walking on inclined surfaces. The results show, that a nonlinear nature of the motor torque, due to a force induced by the slope, gives rise to harmonics of the fundamental walking frequency of 1.73 Hz. These harmonics are the origin of the unstable robot walking. Additionally, the feedback-gain parameters K_A and K_H affect the amplitudes of the harmonics, which give rise to vibrations at a higher surface inclination. Increased surface friction allows a reduction of the feedback gain, which reduces this specific contribution to the harmonics and thus stabilizes the robot. To improve the walking stability on inclined surfaces, it is found that the damped natural frequency of the motor-control system must be kept lower than the fundamental walking frequency.