Optimal design of an asymmetrical parallel mechanism

Abstract

Aiming at the aircraft composite skin grinding, a Three-DOF Asymmetrical Mechanism (TAM) is proposed to replace manual grinding. Considering asymmetrical characteristics of the TAM, the linear superposition principle is adopted to derive the total stiffness matrix of the mechanism. The driving force curves of numerical calculation and simulation are almost coincident; thus the correctness of the dynamic model is verified. The global kinematics condition number index is established with the velocity ellipsoid method. Similarly, the global stiffness performance evaluation index is constructed according to the stiffness ellipsoid method. Moreover, a new global acceleration dexterity index is proposed to overcome the limitations of the dynamics ellipsoid method. Based on the above models and performance indices, a new optimization method is proposed which combines both single and multi-objective optimization. Among the method, the multi-objective optimization is carried out with normalized weighted sum algorithm and genetic algorithm. This optimization method can not only improve the convergence speed, but also balance the weight of different performance indices. After optimization, the kinematics, stiffness and dynamics performance are significantly improved by contrast with the initial performance atlas. Therefore, the results indicate the effectiveness of the multi-objective optimization method.