Time-efficient Windtunnel test method for Obtaining Dynamic Aircraft Characteristics Using Magnetic Suspension and Balance System

Dynamic windtunnel test of standard dynamics model has been conducted using magnetic suspension balance system and identified dynamic derivatives using four different regression methods: linear regression, regression by Bayesian-Information-Criterion (BIC)-optimized model, Ridge and weighted Ridge regression. Dynamic aerodynamic force was described using Gaussian basis function in order to capture continuous and smooth change of the stability derivative against angle of attack (AOA). As a result, static component, pitch rate and acceleration coefficient of longitudinal force components were identified. When compared the regression analysis methods, the BIC-optimized method gave the simplest model and there was little difference among the residual by the regressions below 17-deg AOA. Above 17° the methods suppressed overfitting gave less error for the pitching moment coefficient than those by the linear regression. Z-force by the present method agreed well with the reference value and the pitching moment agreed below 8-deg AOA. Pitching rate derivative of the pitching moment partially agreed in the low AOA range but was larger than those in the reference above 10°.