Simultaneous-Perturbation-Stochastic-Approximation Algorithm for Parachute Parameter Estimation

摘要

This paper presents an algorithm to estimate unknown parameters of parachute models from flight-test data. The algorithm is based on the simultaneous-perturbation-stochastic-approximation method to minimize the prediction error (difference between model output and test data). The algorithm is simple to code and requires only the model output. Analytical gradients are not necessary. The algorithm is used to estimate aerodynamic and apparent mass coefficients for an existing parachute model.