GRAMPC - A gradient based MPC software for real-time applications
GRAMPC is a real-time model predictive control (MPC) tool that is suited for controlling nonlinear systems with input constraints. GRAMPC uses an efficient projected gradient algorithm with an adaptive line search strategy in order to allow an MPC implementation for highly dynamical systems with sampling times in the (sub)millisecond range and/or high-dimensional systems. GRAMPC is implemented in plain C with an additional interface (GUI) to MATLAB/Simulink. A fixed number of gradient iterations is used in each new MPC step and the current solution is used as new initialization (warm-start) to successively refine the predicted MPC trajectories over the single MPC steps.
MIT's spacecraft simulator for control algorithm development
The SPHERES simulation is designed to support most software-related aspects of single- and multi-satellite SPHERES operations. The simulation code base consists of simulated versions of most of the SPHERES core flight code and additional code that simulates dynamics, communications, and other environmental interaction. The simulation is particularly valuable during the early stages of algorithm development and implementation as an aid in accelerating the learning curve for any Guest Scientist for SPHERES. Algorithms may be implemented in C or Embedded MATLAB and executed in the MATLAB simulation environment to verify general desired behavior. With some limitations, the code used in the simulation can be directly transferred to the SPHERES hardware. We are sharing this project with the public to gather more interest for SPHERES Guest Scientists. Feel free to use and contribute to this code base, just don't forget to credit the MIT Space Systems Lab if you distribute any results.
Set of tools for algebraic derivative estimation, based on the work of Fliess and its collaborators.
Simple signal processing projects in Scilab and Matlab
SPHERES Facility software for scientists
BOSAR is a free reinforcement learning blockset for Matlab/Simulink (Mathworks). BOSAR is developed using C/C++ and works with Matlab from 5.3 version. BOSAR is distributed under GNU GPL licence Copyright (c) 2005-2007 FEMTO-ST CNRS.
Motion library for steppers control through PLC
rtmc9s12-Target is a MATLAB/Simulink block-set for real-time control on Freescale MC9S12 microcontrollers.
Symofros is a robotics modeling software originally created by the Canadian Space Agency. Symofros basic functions include modeling Kinematics, Non-linear dynamics, Linear dynamics, Dynamics for constrained systems, and Systems with flexible links.