ESA GNC Conference Papers Repository
Guidance, navigation and control for the autonomous return-to-launch-site functionality of experimental free-falling-units ejected at high altitude from a sounding rocket
Sounding rockets have been a fundamental component of scientific research related to the near-Earth environment. Free-falling units are ejected from high altitudes in order to perform measurements in different regions of the atmosphere and ionosphere. These units perform reentry by deploying a parachute, thus requiring significant time and resources in order to be retrieved, and the recovery operations often happen to be unsuccessful. The BOOMERANG team from the REXUS/BEXUS programme is aiming to address this problem by developing an autonomous return-to-launch-site functionality that can be modularly implemented on free-falling units. This solution makes use of a paraglider that is deployed at a high altitude and autonomously controlled in order to have the units fly to the designated coordinates. The experiment is scheduled to be launched onboard REXUS 32 in March 2024 from Esrange Space Center in northern Sweden. The rocket will reach an apogee of 82 kilometres, while the paraglider will be deployed at an altitude of 10 kilometres. The system will be implemented and launched on disc-shaped units with a mass of 1.7 kilograms and a diameter of 240 millimetres. Multiple approaches to Guidance, Navigation and Control are being investigated and tested to achieve this objective, with different levels of complexity. The hardware that is being used for Navigation purposes includes an accelerometer, a magnetometer, an angular rate sensor and a GNSS receiver. The simplest algorithm that is being considered in order to achieve the goal of the mission makes use of position and velocity in geodetic coordinates from the GNSS receiver to compute the heading error with respect to the target landing coordinates. This error is used to actuate the onboard servo motor that pulls or releases the brake lines of the paraglider, thus steering the flying units. A more complex GNC system is also being implemented and tested. This makes use of a multiplicative extended Kalman filter in order to compute an optimal estimation of attitude quaternions, position and velocity of the flying units. Within the Kalman filter, attitude propagation is performed with the angular rate data, and position and velocity propagation are computed using the accelerometer, while the update uses magnetometer and GNSS data respectively. The precision of estimation is enhanced by the inclusion of the angular rate sensor bias and the accelerometer bias in the navigation code. Information on attitude, position and velocity is used to compute the trajectory to the designated landing location and a PID controller is used in order to provide the required control value for the actuator. The experiment has passed the Critical Design Review and is currently undergoing integration and verification. The system will be validated through a series of flights and drop tests from February to May 2023 that will allow to tune the design.