- The paper describes further developments of the interactive evolutionary design concept relating to the emergence of mutually inclusive regions of high performance design solutions. These solutions are generated from cluster-oriented genetic algorithm (COGAs) output and relate to a number of objectives introduced during the preliminary design of military airframes. The data-mining ofThe paper describes further developments of the interactive evolutionary design concept relating to the emergence of mutually inclusive regions of high performance design solutions. These solutions are generated from cluster-oriented genetic algorithm (COGAs) output and relate to a number of objectives introduced during the preliminary design of military airframes. The data-mining of multi-objective COGA (moCOGA) output further defines these regions through the application of clustering algorithms, data reduction and variable attribute relevance analyses. A number of visual representations of the COGA output projected onto both variable and objective space are presented. The multi-objective output of the COGA is compared to output from a Strength Pareto Evolutionary Algorithm (SPEA-II) to illustrate the manner in which moCOGAs can generate good approximations to Pareto frontiers.…
