Nonlinear Model Predictive Controller Implementation on a Microprocessor for 2D Lunar Powered Descent

Author(s)
Sharma, Saumya
Advisor(s)
Editor(s)
Associated Organization(s)
Organizational Unit
Daniel Guggenheim School of Aerospace Engineering
The Daniel Guggenheim School of Aeronautics was established in 1931, with a name change in 1962 to the School of Aerospace Engineering
Supplementary to:
Abstract
Designing the optimal lunar landing trajectory for a single thrust output, SmallSat form factor vehicle involves analyzing the trajectory from a deorbit burn (braking) and powered descent phase. Due to the need to reduce altitude and velocity while optimizing for maximum landing mass, the system is modeled with non-linear dynamics based on various assumptions which are described in the paper. Open-loop continuous propagation is used with optimal control theory to establish a trajectory that will be used as the source of truth and compared against the hardware implementation of this simulation. Because the designed control algorithm needs to be capable of flying on an on-board computer, a model predictive controller (MPC) was implemented to show how discrete real-time updates impact the optimization of the trajectory. MPC reduces the computational load through an online optimization algorithm instead of using a true optimization problem to produce a more flyable control scheme. To show the effects of running an MPC for a “flight-like” algorithm on a processor that would fit in a SmallSat form-factor, a Raspberry Pi 3B was used to demonstrate how varying the time horizon length and time-step frequency impact computing performance and fuel consumption
Sponsor
Date
2023-05-01
Extent
Resource Type
Text
Resource Subtype
Masters Project
Rights Statement
Unless otherwise noted, all materials are protected under U.S. Copyright Law and all rights are reserved