Concurrent design optimization of the Delta manipulator

The problem of concurrent design of a mechanism can be defined as finding optimal structural parameters and control parameters for a given objective function during the same optimization process. In this paper, a general concurrent optimization methodology for kinematically complex mechanisms is tested using a Delta manipulator. This methodology intends to optimize any structure and control design, using any specified kinematic or dynamic models. Thus, general optimization methods not dependent on mathematical characteristics of the objective function are used. The main contribution of this work is to define, develop and test a general methodology that can generate optimal designs based on workspace and task requirements, such that they guarantee an adequate performance under a set of operating and joint constraints. We test three families of evolutionary algorithms: a genetic algorithm, an evolution strategy and an estimation of distribution algorithm, for two objective functions. The reported results give directions about the most adequate method to tackle the concurrent design problem.