Session
Weekend Session V: Automation - Research and Academia
Location
Utah State University, Logan, UT
Abstract
Attitude state estimation for spacecraft often requires writing an integrator or filter from scratch, with choices depending on the expected sensors, spacecraft and mission properties, and computational resources available. We create a customizable object-oriented satellite dynamics model with a variety of sensors, actuators, and disturbances that can be easily applied to any small spacecraft. Relevant disturbances can be modeled (or ignored), sensor and actuator biases are tracked, and outside effects like eclipse on sun sensors are included. In this work, we demonstrate the use of a new model to create a dynamics-aware unscented Kalman filter (UKF) that, in simulation, outperforms current and previous estimators, even for large satellites and achieves sub-degree accuracy for a small satellite. For example, the mass properties of a satellite, the actuator and sensor properties, and relevant disturbances can be input, and the UKF with relevant dynamics, update, and propagation steps will be created. This allows for rapid testing of various estimation and control paradigms, and the quick development of an attitude determination system (ADS).
Generalized Attitude Estimation for Spacecraft
Utah State University, Logan, UT
Attitude state estimation for spacecraft often requires writing an integrator or filter from scratch, with choices depending on the expected sensors, spacecraft and mission properties, and computational resources available. We create a customizable object-oriented satellite dynamics model with a variety of sensors, actuators, and disturbances that can be easily applied to any small spacecraft. Relevant disturbances can be modeled (or ignored), sensor and actuator biases are tracked, and outside effects like eclipse on sun sensors are included. In this work, we demonstrate the use of a new model to create a dynamics-aware unscented Kalman filter (UKF) that, in simulation, outperforms current and previous estimators, even for large satellites and achieves sub-degree accuracy for a small satellite. For example, the mass properties of a satellite, the actuator and sensor properties, and relevant disturbances can be input, and the UKF with relevant dynamics, update, and propagation steps will be created. This allows for rapid testing of various estimation and control paradigms, and the quick development of an attitude determination system (ADS).
