作者: Jong-Woo Kim , Roberto Cristi , Brij Agrawal
DOI: 10.2514/6.2004-5386
关键词: Linearization 、 Simulation 、 Sun sensor 、 Kalman filter 、 Gyroscope 、 Extended Kalman filter 、 Filter (signal processing) 、 Spacecraft 、 Engineering 、 Control theory 、 Quaternion
摘要: This paper presents the attitude determination method for Bifocal Relay Mirror Spacecraft Simulator. The simulator simulates three-axis motion of a spacecraft and has an optical system emulating bifocal space telescope. consists three control moment gyroscopes, rate gyros, two-axis analog sun sensor, two inclinometers. five-foot diameter platform is supported on spherical air bearing to offer low-torque environment. demonstrates methods employing measurements from IR inclinometers, triaxial gyroscope. first implements conventional Kalman filter algorithm. second uses nonlinear observer derived Lyapunov's direct method. Analytical experimental results are presented validate proposed TASS simulating single five- foot low- torque variable speed gyroscopes as actuators, gyros angular rate, sensor inclinometers sensors. Detailed description hardware given in Ref. 2. focuses development validation algorithm using Although cannot be used spacecraft, they can still employed testbeds get useful information with good accuracy at low cost. Also, if light source not far enough, effect transitional during rotation testbed should taken into account when calibrating sensor. Quaternion parametrization offers singularity-free represent three-dimensional rota- tions four-dimensional unit vectors. But four elements quaternion estimated independently because normality constraint. It standard approach include gyro bias overall kinematics model estimate actual by Extended Filter (EKF) techniques. EKF, extension (KF) models, works satisfactorily linearized exhibits sufficient accuracy. Since EKF utilizes linearization each step, constraint quaternions rigidly enforced within computational structure. Various have been developed overcome this deficiency described Refs. 3, 4, 5, 6. nominal basic six-state estimation known Multiplicative (MEKF). 4 In MEKF formulation TASS, Modified Ro- drigues Parameters (MRPs) unconstrained representation error