Component Based Modelling And Optimization For Modular Robot Design

In this paper, an approach for modular design of industrial robots is presented. The approach is to introduce an objectoriented simulation model of the robot and combine this with a discrete optimization algorithm. The simulation model of the industrial robot is developed in Modelica, an object oriented modeling and simulation language, and simulated in the Dymola tool. The optimization algorithm used is a modification of the Complex method that has been developed in Matlab and connected to the simulation program. The optimization problem includes selecting components such as gearboxes and motors from a component catalogue and the objective function considers minimization of cost with constraints on gear box lifetime. Furthermore, the correctness of the model has been verified by comparison with an in-house simulation code with high accuracy.