Optimal Design for Biped Robot Parallel Ankle Joint

Abstract

After analyzing the motion mechanism of human ankle, a new type of parallel ankle joint mechanism for a biped robot is designed. The parameterized model of the mechanism is established and its kinematic and dynamic analyses are carried out, with the conclusions of the first-order influence matrix of the parallel mechanism and the relationship between the driving force in the concertina motion and the zero moment point trajectory. Considering the great difference between the actual angular velocity and the driving moment in the biped robot’s walking ankle joint, the structure parameters of the parallel mechanism are optimized through iterative calculation according to the dynamic characteristics of the ankle joint during walking, which make the driving peak power and speed of the two driving elements of the ankle joint tend to coincide and reduce the dynamic unbalance in the biped robot’s walking ankle joint. According to the analysis of the experimental results of the biped robot’s ankle joint motion, the peak-power used to drive the optimized parallel ankle joints is only about half of that used to drive serial joints, which lowers the demand of the design power for the driving components. On the premise that the dynamic characteristic of biped robot ankle joints during walking are satisfied, choosing driving components with lower power is beneficial to the reduction of bulk, mass and the energy consumption of the joints.