Bringing rockets safely back to the ground

Have you ever watched along as the SpaceX "Falcon" returns to land and wondered how it is possible for a rocket to take itself from flight to a vertical landing? What factors need to be considered? How do you build a control system that can cope with unforeseen weather?
Thrust vector controlled approach
Thrust vector test drone: Full view
Thrust Vector Control (TVC) test platform.

It is exactly these kinds of questions that the researchers of the EPFL Rocket Team are trying to answer with their test platforms. The team use a dynamic Model Predictive Controller (MPC) and plug in their control, navigation and guidance algorithms to enable to robot to follow the best trajectory towards its landing goal. By using predictive control coupled with real-world data, it is possible to estimate the position and trajectory of the robot and to therefore optimise the control instructions.

By having this continuously updating control, it is possible for the robot to adapt to current needs during flight; meaning that it can either use the least fuel possible, thus reducing the environmental impact, or when a storm approaches, the robot can abandon previous commands and land as quickly as possible.

Have a look at the EPFL Rocket Team at https://epflrocketteam.ch/ and a video on the TVC test rocket on YouTube.

 

We provide an inexpensive test platform to develop a new generation of guidance, navigation and control algorithms.

Contact

Roland Schwan

Roland Schwan

roland.schwan@epfl.ch
Johannes Waibel

Johannes Waibel

johannes.waibel@epfl.ch