Classification of Aircraft Design Variables using Principal Component Analysis in Aircraft Conceptual Design

Abstract

It is sometimes difficult for aircraft designers to grasp the overall picture during the conceptual design process because the process is highly complicated. Various parameters related to an aircraft, such as design requirements and design variables, entangle each other with complexities. This paper aims to evaluate whether Principal Component Analysis (PCA), which is used to analyze multidimensional data, is applicable to clarify the relationships among design requirements and design variables when a conceptual design example is considered. First, an aircraft conceptual design tool, using a multi-objective genetic algorithm, is constructed. The validity of this tool is confirmed by comparing the design results with an existing aircraft. The design tool is used to create a data set of aircraft. From the data set, five important parameters are chosen and PCA is applied to them. They are classified into two groups based on their characteristics, and the relationships among them are discussed. The results indicate that PCA could be used to provide beneficial information for the aircraft conceptual designers.