Abstract: In order to improve the driving performance of existing robot,an artificial-muscle-based method is proposed to design novel driving joints.Firstly,the Cartesian coordinate system of the joint mechanism is defined,and a mathematical model is constructed to express the relationships among the structure parameters,workspace,structure intensity,and dynamical features of the driving joints.Based on these analyses,a multi-objective programming algorithm is then utilized to optimize the design of the driving joints,and such relevant factors as the design variables,objective function,constraint conditions and solution method are presented in detail.Finally,the proposed method is applied to the sciatic joint design of a quadruped robot,and the simulation results prove the advantages of the artificial-muscle-based driving joint in terms of strength,dexterity, and carrying capacity.