Hybrid Toolbox for Matlab
|
The Hybrid Toolbox is a Matlab/Simulink toolbox for modeling, simulating, and verifying hybrid dynamical systems, for designing and simulating model predictive controllers for hybrid systems subject to constraints, and for generating linear and hybrid MPC control laws in piecewise affine form that can be directly embedded as C-code in real-time applications. |
|
The main features of the toolbox are:
|
Hybrid model design and simulation Mixed logical dynamical (MLD) systems are handled as Matlab objects, obtained from HYSDEL models. MLD objects can be automatically converted into piecewise affine (PWA) objects. MLD and PWA objects can be simulated in Matlab or in Simulink. Safety properties can be verified through reachability analysis. |
|
|
Control design Model predictive controllers based on on-line optimization (MILP / MIQP) can be designed for hybrid systems. Quadratic and infinity norms are supported. |
|
|
Explicit control design Piecewise affine controllers can be designed via offline optimization (multiparametric programming) for linear systems w/ constraints and for hybrid systems. MPC controllers designed for linear constrained systems with the new Model Predictive Control Toolbox for Matlab (v.2.1) can be also converted into piecewise affine form via multiparametric quadratic programming. |
|
|
C-code generation Explicit controllers can be easily exported as C-code for direct implementation and rapid prototyping, by either simply running Real-Time Workshop or by embedding the generated C code in the application. |
|
|
Simulink Library Blocks are provided for: controllers based on on-line optimization, explicit piecewise linear controllers, MLD and PWA models. |
|
|
Solver Support The Hybrid Toolbox supports several solvers for linear, quadratic, and mixed-integer optimization. |
List of supported solvers:
|
Code/Manual Download (Version 1.2.3 - March 5, 2010)
Note: The code can be
downloaded only for non-commercial use at universities, governmental
entities, or academic non-profit institutions. An academic non-profit institution is defined by its
mission: research in the public domain and education. A non-profit
organization that performs research as a service under contract to a
commercial entity does not belong to the category of non-commercial
users. Commercial users interested in obtaining the code should contact
the author Alberto Bemporad.
Citations to the toolbox manual are appreciated:
@MISC{HybTBX,
author = {A. Bemporad},
title = {{Hybrid Toolbox - User's Guide}},
note={\url{http://www.ing.unitn.it/~bemporad/hybrid/toolbox}},
year = {2004},
}
Toolbox download:
>>> Download request #2800 received on December 22, 2009 <<<
Please choose your platform below:
|
|
|
|
|
|
|
|
|
|
|
Author: Alberto Bemporad
Note: The Model Predictive Control Toolbox for Matlab should be installed to use the Hybrid Toolbox. For simulation of hybrid MPC controllers based on mixed-integer optimization, CPLEX or Xpress-MP are highly recommended.
You are visitor number
|
Acknowledgements: the Scientific Research Laboratory of Ford Motor Company (Dearborn, MI) is acknowledged for continuous support. The European Commission is also acknowledged for supporting several theoretical developments through EU project "CC" (IST-2001-33520). |
||
|
|
|
|
(C) 2003-2010 by A. Bemporad
