Rendezvous & Proximity Operations

This respository contains the code for the restricted two-body simulator and controllers that were implemented for the final project of my Spring 2022 ME 601 - Autonomous Feedback course.

The final report can be found here.

Example Trajectory Tracking Simulation

This simulation tracked a box trajectory around the target in a circular orbit at an SMA of approx 30,000 km. using the time varying LQR trajectory tracking controller.

Structure

The proposal/ and report/ directories contain the documents that were actually turned in. The simulator/ directory contains the source code.

Simulator Architecture

There is a Simulator class that takes in Vehicle objects, and propogates the dynamics of the system for a given duration. Each Vehicle can define its own control law.

Each controller has it's own source file, and each one inherits from the toplevel Vehicle class. The four controllers are can be found here:

• Infinite LQR static state feedback controller
• Time invariant trajectory tracking LQR controller
• Time varying trajectory tracking LQR controller
• Time invariant MPC controller

There is one primary executable (rendezvousMain), that can be used to run and visualize a set of different simulations, each of which are defined as a function in SimulationRuns.cpp.

Building

Eigen3 is the only dependency that won't be downloaded at configure time. To build, simply clone and use the provided utility script:

$ git clone git@github.com:ianruh/Rendezvous-Proximity.git
$ cd Rendezvous-Proximity/simulator
$ ./utils build

or use CMake directly:

$ git clone git@github.com:ianruh/Rendezvous-Proximity.git
$ cd Rendezvous-Proximity/simulator
$ mkdir build/ && cd build/
$ cmake ..
$ make

Running Simulations

Once a simulation run has been defined in the SimulationRuns.cpp file a registered in the map of simulations at the top of main.cpp, then it can be run using the rendevousMain binary directly:

Usage: rendevousMain [options] action input output_name

Positional arguments:
action          Either 'simulate' or 'visualize'.
    Example: rendevousMain simulate below200InfiniteLQR below200output
    Example: rendevousMain visualize below200output.csv below200video

input           Either a simulation name or a record csv.
output_name     The output name (no final extension).

Optional arguments:
-h --help       shows help message and exits [default: false]
-v --version    prints version information and exits [default: false]

Or you can use the utils helper script (this is recommended, as I put minimal effort into the usability of the binary's CLI):

$ ./utils
Usage: utils <subcommand> [options]\n
Subcommands:
    build                 Build in release
    build-debug           Build in debug
    clean                 Clean build artifeacts
    run [name]            Run simulation and generate outputs
    runo [name]           Run simulation, generate and open outputs

For help with each subcommand run:
utils <subcommand> -h|--help

$ ./utils run below200InfiniteLQR
Simulating and visualizing below200InfiniteLQR finished.

Dependencies

• cppmpc for the MPC controller.
• SymEngine for the symbolic mathematics.
• Eigen3 for linear algebra and basic vector/matrix types.
• Matplot++ for plotting and visualization.
• rapidcsv for CSV reading.