Abstract
Topology optimization (TO) is a promising field for the design of industrial devices, since it can allow computational tools to develop creative solutions independently of initial topologies. However, the generation of smooth solutions, i.e., solutions that can be physically implemented, remains a challenge for most TO techniques available. In this paper, we discuss a set of operators that allow smoother topologies to be generated by an evolutionary topology optimization algorithm based on the artificial immune systems paradigm. The design of an electromagnetic device is used as a test case for the proposed methodology.