Data mining for multidisciplinary design space of regional-jet wing

Two data mining techniques have been performed for a large-scale and real-world multidisciplinary design optimization (MDO) results to provide knowledge in design space. The MDO among aerodynamics, structures, and aeroelasticity for a regional-jet wing was carried out using high-fidelity evaluation tools on adaptive range multi-objective genetic algorithm. MDO generated 130 solutions included in nine non-dominated solutions. All solutions were investigated regarding tradeoffs among three objective functions and effects of design variables on objective functions using a self-organizing map (SOM) and a functional analysis of variance (ANOVA) to extract key features of the design space. Consequently, as SOM and ANOVA compensated with respective disadvantages, the design knowledge could be obtained more clearly by the combination between them. Although the MDO result showed the inverted gull-wings as non-dominated solutions, one of the key features revealed by data mining was the non-gull wing geometry. When this knowledge was applied to one optimum solution, the resulting design was found to have better performance compared with the original geometry designed in the conventional manner. Data mining can discover better design due to the salvage of information from design space even when optimization itself does not converge.