A Seven Link Biped Robot Walking Pattern Generation on Various Surfaces

Abstract

This Paper presents a periodic approach with scalable walking trajectory based on gait characteristic parameters including length, maximum height and time cycle. The proposed methodology is split in two parts: robot trajectory and dynamic stability examination. The lower body is in charge of general bipedal walking trajectory where a limited number of breakpoints in both stable and unstable phases are identified. Consequently, three joints (hip, knee and ankle) positions are derived for a seven link biped robot. Considering the fields of Computational Intelligence and Soft Computing leads to an efficient non-conventional approach to generate a smooth walking trajectory. The methodology for walking pattern planning based on Artificial Neural Networks using Radial Basis Function intends to fit a curve on derived breakpoints. The biped robot stability during walking cycles is investigated by the Zero Moment Point (ZMP) criterion. For the dynamic stability study, the ZMP for a stable condition in a determined polygon of support in every single gait step is calculated. For trunk motion adjustment and lower limb movement compensation, a Linear Inverted Pendulum model and ZMP criterion are employed to obtain upper body trajectory satisfying whole robot walking dynamic stability.