Kalman Filter Application for Flexible Space System Parameters Identification

Abstract

Nowadays an increasing number of space missions are using satellites with a rigid hub and long and/or wide flexible appendices, such as solar panels, communication antennas, telescopic structures, robotics flexible arms. The increasing need for better pointing accuracy of antennas connected to the rigid part of the spacecraft leads to a requirement for more efficient controllers, where more accurate identification process play an important rule into the closed loop system. To meet the requirements for pointing accuracy, flexible parameters as elastic displacement, which are of great importance for control tasks should be continuously identified in the space environment. This paper presents an investigation results of elastic displacement identification using the Kalman Filter methodology. A flexible Euler-Bernoulli beam, connected to a rigid core with torques as input and angles and angular velocities as outputs is used as simple mathematical model of a rigid-flexible satellite to apply the Kalman filter identification algorithm proposed. The Kalman filter is tested under several conditions considering cold start with coarse initial knowledge and varied measurement noise levels. At the end comments are drawn about the robustness of the proposed procedure and feasibility of implementation within the control system loop.