Topology Optimization of Robots Using Mapping Techniques

This work is an extension of previous work that utilizes mapping techniques to handle the discrete nature of robot design. The developed design approach utilizes dimensionless parameters that map data bases of commercially available components. The important properties are derived by applying interpolation on the data bases. The handling of both distinct components sub-groups as well as a method for the gradual conversion of the dimen-sionless variables into integers has already been presented. In this work the approach is further developed by introducing a parameter that maps a set of possible system topologies. The possible topologies include three different spatial mechanisms and the dynamic time domain simulation model that is used to evaluate a design is developed so that it can analyze each of the possible topologies. The mapping of the robot topologies has been done in such a way that the performance of a design varies continuously despite a change in topology during the optimization. The design criteria include costs of drives and structural components, tool point precision, fatigue in welded details, over heating and stalling of the motors and gears as well time of operation. A number of examples are given.