A novel 6 DOFs generalized parallel manipulator design and analysis based on humanoid leg

Abstract

In this paper, a 6 degrees of freedom (DOFs) generalized parallel manipulator is presented which can be used as a humanoid leg. Based on the Number Synthesis and Graph Synthesis method, all the possible contracted graphs of the 6-DOFs leg are provided. Then, a novel method named as Node Response Method (NRM) is presented to identify the isomorphism of contracted graphs. In order to obtain the optimal manipulator, the Lie group is used to analyze all limbs. Furthermore, a terrain adaptive index named as Global Maximum Stride Distance Ratio (GMSDR) is established, which is used to assess the performance of a biped robot in different terrains. The DOF of proposed manipulator is verified by translating into a two-layer and two-loop (TL-TL) topological structure. Afterwards, the kinematics and the Jacobian matrix of the manipulator are calculated. Moreover, the performance indexes of the manipulator are analyzed. In order to obtain an optimal kinematics performance, the structural parameters are optimized using genetic algorithm. Finally, two basic gaits are proposed for a biped robot which consists of two manipulators and a simulation is carried out.