Attitude estimation and control based on modified unscented Kalman filter for gyro-less satellite with faulty sensors

Abstract

A modified unscented Kalman filter is presented to estimate the quaternion parameters as well as the angular velocities of a rigid gyro-less satellite under faulty sensor conditions. The task is carried out using the Sun sensor and magnetometers as attitude sensors with bounded noise and unknown fault(s). Following the presentation of the satellite attitude dynamics and filtering formulations, a new fault detection and isolation algorithm is proposed. The latter is based on a modified unscented Kalman filter structure for improved fault detection, sensor isolation, and attitude control (AC). A Backtracking Search Algorithm (BSA) is also used to design and optimize the PID controller gains, as well as to optimize sensor installation orientation for residual signal decoupling. In this regard, the paper presents and proves a theory that describes how the proposed optimization process for sensor installation can analytically decouple the faulty orientation signal from the healthy ones. A comparison of attitude and angular velocity estimation approaches using the standard Kalman, extended Kalman, and a proposed modified unscented Kalman filter is also provided. The results show that the proposed filter performs significantly better in closed-loop AC systems and copes adequately with sensor faults.