On estimation of vehicle linear model parameters

摘要

Purpose - The purpose of this paper is to identify linear model parameters of launch vehicles based on the actual flight test data. To compare the estimated parameters with the ones obtained by two other approaches: identification based on the recorded data from six-degree-freedom simulation of motion and linearization of the equations of motion via small-disturbance theory as an analytical method. Design/methodology/approach - As the vehicle contains all the key issues in system identification such as time-varying, unstable, nonlinear, and closed-loop dynamics, Kalman filter method under the autoregressive with exogenous input model structure is used as a powerful method to estimate the dynamic parameters. Findings - Simulation results demonstrate that the linear model parameters used in the vehicle design and analysis should be validated by flight test data to accurate the vehicle dynamic model as more as possible. Practical implications - One of the most important usages of a linear model of aerospace vehicles is to design their controller. Another application of the algorithm presented in this paper is to estimate online dynamic parameters of the vehicle when they are required for the operation of the control system. Originality/value - Being strongly affected by vehicle dynamic characteristics, linear model parameters of launch vehicles play important part in their design and analysis.