Parallel evolutionary algorithms in shape optimization of heat radiators

The paper deals with the application of Parallel Evolutionary Algorithms (PEA) and the Finite Element Method (FEM) in shape optimization of heat radiators. The fitness function is computed with the use of the coupled thermoelsticity modelled by MARC/MENTAT software. The geometry, mesh and boundary conditions are created on the basis of a script language implemented in MENTAT. In order to reduce the number of design parameters in evolutionary algorithms, the shape of the structure is modelled by Bezier curves. Numerical examples for some shape optimization are included.