Optimal design of a parallel assembling robot with large payload-to-mass ratio

Abstract

• Kinematic, stiffness and mass modeling of a parallel assembling robot is conducted. • A meta modeling method is proposed to consider topological variation of the base. • Optimal design towards large payload-to mass ratio is carried out. • Different objective arrangements are discussed based on Pareto-based optimizations. Assembly in a confined space, such as the cabin of an aircraft or train, demands the assembling device to be with compact structure, satisfactory kinematics and excellent load carrying capability. A six degree-of-freedom (DoF) parallel robot is proposed and designed for such assembling tasks in this paper. Specially, internal structure changes introduced by topology optimization are considered and the multi-objective optimization reaching large load-to-mass ratio is implemented. First, external dimensions of the base and the ratio of the remaining volume to the complete volume are the inputs. Once a set of input variables are given, the topology optimization will be performed by FEA software to form the structure of the base. The stiffness and mass of the base, being the outputs, are obtained numerically by software. Then, the meta models are established by the response surface model (RSM) method. On this basis, stiffness and mass models of the robot are built by the semi-analytical method. The optimal design is implemented by Pareto-based multi-objective optimization. Different arrangements of the objectives are compared. The results show that kinematic indices on the Pareto fronts are all at a satisfactory level. The optimization having payload-to-mass ratio as objective leads to the optimum with higher stiffness along z -axis and smaller mass. The design 6-DoF parallel assembly robot can carry up to 42.06 kg loads while the mass is only 12.002 kg.