PREDICTION OF RESPONSE VIBRATION FROM LONGITUDINAL DATA ANALYSIS BASED UPON A SMALL NUMBER OF INDEPENDENT VARIABLES

摘要

The availability of measured independent descriptor functions of time (e.g., aircraft Mach Number (M), normal acceleration （N{sub}z）, dynamic q (q), Angle-of-Attack (α)) allows for prediction of dependent descriptor variables in time, (e.g., mechanical response acceleration) by way of the Generalized Linear Hypothesis (GLH). For an ensemble of similar but independent measurement time records the GLH formulation can be enhanced by considering a "longitudinal data analysis"(LDA) formulation. LDA is a modeling framework that has been successfully applied both in the social and biological sciences to time history data. This paper provides the basis for analysis of fifteen measured nonstationary acceleration vibration time records using the GLH in conjunction with LDA. It is noted that this technique provides complementary information to the "cross-sectional" nonlinear regression analysis that has been used in the past to predict vibration levels on captive carry stores given average aircraft parameters of q and α. This paper also provides insight into the relationship between nonstationary vibration response analysis in the frequency domain and variation in the descriptor functions of time.