MPC-BSpline

This repository provides a Python-based simulator designed to implement Model Predictive Control (MPC) for B-Spline trajectory tracking. The simulator is built to handle trajectory tracking for an omnidirectional robot, offering a flexible and efficient way to test and tune control parameters.

Installation

Clone this repository:

git clone https://github.com/C2SR/MPC-BSpline.git
cd MPC-BSpline
pip install -r requirements.txt

Usage

Use the trajectory_generator.py script to create B-Spline trajectories. This script accepts various command-line arguments to configure path generation parameters:

trajectory_generator.py [-h] [--map MAP] [--k K] [--Lambda LAMBDA] [--vmax VMAX] [--vmin VMIN] [--delta DELTA] [--path PATH] [--f F] [--file FILE] [--folder FOLDER]

Available Arguments

Argument	Description
-h, --help	Show the help message and exit.
--map	Path to the map YAML file.
--k	B-Spline order.
--Lambda	Corner detection hyperparameter for B-Spline.
--vmax	Maximum speed for the robot.
--vmin	Minimum speed for the robot.
--delta	Acceleration parameter for the robot.
--path	B-Spline control points for the trajectory.
--f	Control cycle frequency.
--file	Name of the output file.
--folder	Directory where the output file will be saved.

To use the trajectory specified in the associated paper, run the script with default values. The script generates trajectories in both .npy and .yaml formats—.npy for this repository and .yaml for use with ROS.

The figure below illustrates the control point IDs used in the B-Spline path:

Other maps may be used, provided they are represented as bidirectional graphs with nodes and edges stored in a YAML file.

Running the MPC Simulator

To start the MPC simulator, execute:

python3 main.py --cfg config_omni.yaml

This configuration file includes:

• Simulation Parameters: Defines the simulation length, control step size, and physical dimensions.
• Motor Parameters: Configures motor characteristics, including gain, time constant, and control limits. The motor is implemented as a first order system.
• MPC Parameters: Specifies MPC tuning variables such as the prediction horizon (N) and weighting matrices (Q and R).

Dependencies

This project requires Python 3.x and additional libraries listed in requirements.txt.

License

This project is licensed under the MIT License. See the LICENSE file for details.