Welcome to nMPyC’s documentation!

nMPyC is a Python library for solving optimal control problems via model predictive control (MPC).

nMPyC can be understood as a blackbox method. The user can only enter the desired optimal control problem without having much knowledge of the theory of model predictive control or its implementation in Python. Nevertheless, for an advanced user, there is the possibility to adjust all parameters.

This library supports a variety of discretization methods and optimizers including CasADi and SciPy solvers.

In summary, nMPyC
  • solves nonlinear finite horizon optimal control problems

  • solves nonlinear optimal control problems with model predicitve control (MPC)

  • uses algorithmic differentation via CasADi

  • can chose between different discretization methods

  • can chose between different solvers for nonlinear optimization (depending on the problem)

  • supports time-varying optimal control problems

  • supports the special structure of linear-quadratic optimal control problems

  • supports discounted optimal control problems

  • can save and load the simulation results

The nMPyC software is Python based and works therefore on any OS with a Python distribution (for more precise requiremnents see the Installation section). nMPyC has been developed by Jonas Schießl and Lisa Krügel under the supervision of Prof. Lars Grüne at the Chair of Applied Mathematic of University of Bayreuth. nMPyC is a further devolpement in Python of the Matlab code that was implemented for the NMPC Book from Lars Grüne and Jürgen Pannek [GruneP17].