State Estimation and Control of a Self-Foraging Rocket

Tilstands Estimering og Regulering af Raket

Hemme, Kasper ; Nauheimer, Michael

4. term

Control and Automation, Master

2015

2015-06-03

162

Dette speciale undersøger, hvordan man bestemmer og styrer en rakets position og orientering (attitude) med fokus på en billig og effektiv løsning til et voksende mikrosatellitmarked. Arbejdet tager udgangspunkt i 'self-foraging rocket'-konceptet, inspireret af Yemets et al. ved Dniepropetrovsk National University. Raketten designes fra grunden: det indledende mekaniske koncept, valg af sensorer (der måler bevægelse og orientering) og aktuatorer (dele der kan styre/bevæge raketten), samt dimensionering af hver thruster (små raketdyser). Der beregnes en generel opstigningstrajektorie til baneindsættelse og den testes i en komplet simuleringsplatform i Simulink. Et udvidet Kalman-filter (EKF) udvikles til at estimere rakettens tilstande, som fx position, hastighed og orientering, ud fra sensordata. Til styring implementeres to tilgange: en waypoint-følgende regulator, der holder raketten på en optimal opstigningsrute, og en sliding-mode-styring, en robust kontrolmetode. I tre nominelle tests gav EKF’et tilstrækkeligt nøjagtige estimater. Begge styringsmetoder viste god reguleringspræstation, men ingen af dem formåede at indsætte raketten i bane med tilfredsstillende resultat i simuleringen. Specialet præsenterer dermed en samlet design- og simuleringsramme og peger på, hvad der skal forbedres for at opnå pålidelig baneindsættelse.

This thesis investigates how to determine and control a rocket’s position and orientation (attitude) with an emphasis on a low-cost, efficient solution for the growing microsatellite market. It explores the 'self-foraging rocket' concept, inspired by Yemets et al. at Dniepropetrovsk National University. The rocket is designed from the ground up: the initial mechanical concept, selection of sensors (to measure motion and orientation) and actuators (devices that steer/move the rocket), and the sizing of each thruster. An ascent trajectory for orbit insertion is computed and tested in a complete simulation environment built in Simulink. An Extended Kalman Filter (EKF) is developed to estimate the rocket’s states—such as position, velocity, and attitude—from sensor data. Two control approaches are implemented: a waypoint-tracking controller to follow an optimal ascent path, and a sliding mode controller, a robust control method. In three nominal test cases, the EKF produced sufficiently accurate estimates. Both controllers showed good control performance, but neither achieved satisfactory orbit injection in simulation. The work presents an end-to-end design and simulation framework and highlights what needs improvement to achieve reliable orbit insertion.

[This abstract was generated with the help of AI]

Space ; Raket ; Regulering ; Autmation ; Estimering ; Optimering ; Missil

Download
View record in AAU Student Projects