Foot Placement Estimator for stepping down movement

In this paper, a novel methodology is proposed for a biped(humanoid) robot to step down stairs using FPE(Foot Placement Estimator)[1]. Firstly, we defined a biped model as 7 mass model and various geometric and inertial properties are defined for inducing the initial angular velocity of the biped robot. We divided the stepping down motion as 4 consecutive motions. The objective of the proposed foot placement estimator is for the biped robot not to falling down during stepping down and make the final kinematic energy to zero value by estimating the initial angular velocity of the biped robot. The numerical simulations are performed to verify the proposed methodology. The results shows that the initial angular velocity of the biped robot decreases as the stepping angle(the angle between the current stair and the next stair) is increased and increases as the stepping distance (the distance between the current stair and the next stair) increases. There are stepping distances at which the initial angular velocity is zero for a specified stepping angle and stepping distance.