A Simple Feedback Controller to Reduce Angular Momentum in ZMP-Based Gaits

Abstract

The Zero Moment Point (ZMP) stability criterion has been broadly employed for walking pattern generation in legged robots. However, ZMP‐based approaches usually ignore the presence of angular momentum in the system. This hinders the performance of the gait, especially against disturbances. In this work we propose an angular momentum controller that can be integrated into standard ZMP‐based gaits. The experiments on a real Nao robot demonstrate that the use of the proposed angular momentum controller improves the stability of a walking pattern generator based on the preview control of the ZMP

An adaptive neuro-fuzzy based on a fractional-order proportional integral derivative design for a two-legged robot with an improved swarm algorithm

“…The foot position of the TLR represents the balance of the robot around a certain trajectory. The concept used to measure the foot balance of the TLR is called zero moment point (ZMP) [22], [36]. The same procedure is used for the stairs descending, where Figure 13 present the stairs descending for TLR while the obtained trajectories for the ANFFOPID controller and the MCIWO-NN are shown in Figure 14.…”

“…Related to this work, many researchers across the globe present and develop different controllers such as [18][19][20], where the authors proposed on gait generation for a 7-DOF robot. While in [21,22], they present a PID controller with ZMP to test the various joints of motors.…”