Knowledge discovery in multidisciplinary design space for regional-jet wings using data mining

Data mining is an important facet of solving multi-objective optimization problems. In the present study, two data mining techniques were applied to a large-scale, real-world multidisciplinary design optimization (MDO) problem to provide knowledge regarding the design space. The use of MDO in the aerodynamics, structure, and aeroelasticity of a regional-jet wing was carried out using high-fidelity evaluation models with an adaptive range multi-objective genetic algorithm. As a result, nine non-dominated solutions were generated and used for tradeoff analysis of three objectives. All solutions evaluated during the evolution were analyzed for the influence of design variables using a self-organizing map (SOM) and a functional analysis of variance (ANOVA) to extract key features of the design space. As SOM and ANOVA compensate for respective disadvantages, the design knowledge could be obtained more clearly by combinating them. Although the MDO results showed inverted gull-wings as non-dominated solutions, one of the key features found by data mining was a 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.