Nonlinear Progamming with CasADi
Project description
Nonlinear Programming with CasADi
csnlp provides classes and utilities to model, solve and analyse nonlinear programmes (NLPs) in optimization.
In particular, it makes use of the CasADi framework [1] to model the optimization problems and perform symbolic differentiation, as well as the IPOPT solver [2] (though the package can be adapted to other solvers pretty easily). The package offers also tools for the sensitivity analysis of NLPs, solving them with multiple initial conditions, as well as for building MPC controllers.
Installation
To install the package, run
pip install csnlp
csnlp has the following dependencies
For playing around with the source code instead, run
git clone https://github.com/FilippoAiraldi/casadi-nlp.git
Usage
Similar to CasADi Opti, we instantiate a class that represents the NLP and allows to model its variables, parameters, constraints and objective. A simple example is below:
from csnlp import Nlp
nlp = Nlp()
x = nlp.variable("x")[0]
y = nlp.variable("y")[0]
p = nlp.parameter("p")
nlp.minimize((1 - x) ** 2 + 0.2 * (y - x**2) ** 2)
g = (x + 0.5) ** 2 + y**2
nlp.constraint("c1", (p / 2) ** 2, "<=", g)
nlp.constraint("c2", g, "<=", p**2)
nlp.init_solver() # initializes IPOPT under the hood
sol = nlp.solve(pars={"p": 1.25}) # solves the NLP for parameter p=1.25
x_opt = sol.vals["x"]
y_opt = sol.value(y)
The package also allows to enhance the NLP with different capabilities with, e.g., multistart (see csnlp.MultistartNlp
) or by wrapping it. As of now, wrappers have been implemented for
- sensitivity analysis (see
csnlp.wrappers.NlpSensitivity
; based on [3]) - Model Predictive Control (see
csnlp.wrappers.Mpc
; [4] for details) - NLP scaling (see
csnlp.wrappers.NlpScaling
).
For instance, to compute the sensitivity of the optimal primal-dual variables y
with respect to the parameters p
of the NLP , first we need to augment the capabilities of the NLP with a wrapper specialized in differentiating the optimization problem, and then compute the first-order and second sensitivities (dydp
and d2ydp2
, respectively) as such:
from csnlp import wrappers
nlp = wrappers.NlpSensitivity(nlp)
dydp, d2ydp2 = nlp.parametric_sensitivity()
In other words, these sensitivities provide the jacobian and hessian that locally approximate the solution w.r.t. the parameters p
. As shown in the examples, the sensitivity can be also computed for any generic expression z
that is a function of the primal-dual variables and the parameters, i.e., z(x(p),lam(p),p)
, and computations can be carried out symbolically or numerically (more stable).
Similarly, an NLP instance can be wrapped in an MPC wrapper that makes it easier to build such controller.
Examples
Our examples subdirectory contains example applications of this package in NLP optimization, sensitivity analysis, scaling of NLPs, and optimal control.
License
The repository is provided under the MIT License. See the LICENSE file included with this repository.
Author
Filippo Airaldi, PhD Candidate [f.airaldi@tudelft.nl | filippoairaldi@gmail.com]
Delft Center for Systems and Control in Delft University of Technology
Copyright (c) 2023 Filippo Airaldi.
Copyright notice: Technische Universiteit Delft hereby disclaims all copyright interest in the program “csnlp” (Nonlinear Progamming with CasADi) written by the Author(s). Prof. Dr. Ir. Fred van Keulen, Dean of 3mE.
References
[1] Andersson, J.A.E., Gillis, J., Horn, G., Rawlings, J.B., and Diehl, M. (2019). CasADi: a software framework for nonlinear optimization and optimal control. Mathematical Programming Computation, 11(1), 1–36.
[2] Wachter, A. and Biegler, L.T. (2006). On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming. Mathematical Programming, 106(1), 25–57.
[3] Büskens, C. and Maurer, H. (2001). Sensitivity analysis and real-time optimization of parametric nonlinear programming problems. In M. Grötschel, S.O. Krumke, and J. Rambau (eds.), Online Optimization of Large Scale Systems, 3–16. Springer, Berlin, Heidelberg
[4] Rawlings, J.B., Mayne, D.Q. and Diehl, M., 2017. Model Predictive Control: theory, computation, and design (Vol. 2). Madison, WI: Nob Hill Publishing.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.