A novel two-step parameter identification method for GyroWheel system

Abstract

GyroWheel is a rate sensing/momentum management instrument for small spacecrafts. The challenge to realize accurate angular rate estimation consists in the identification of system parameters, which is investigated in this paper. An inverse dynamic model is derived by ignoring the high-frequency components, and it is the foundation of parameter identification. Then a two-step method is developed to solve the ill-conditioned problem of parameter identification. In the first step, the inverse dynamic identification model is simplified in some special situations, and a rough estimation of the parameters can be obtained. In the second step, a constrained optimization problem is obtained by introducing the results of the first step, and particle swarm optimization algorithm is utilized to realize the identification of these parameters. Validation simulations and experiments on a GyroWheel prototype show that the proposed method is a good promising method for offline parameter identification.