acados¶
Fast and embedded solvers for nonlinear optimal control.
acadossource code is hosted on Github. Contributions via pull requests are welcome!acadoshas a discourse based forum.acadosis mainly developed by the syscop group around Prof. Moritz Diehl, the Systems Control and Optimization Laboratory, at the University of Freiburg.
About acados¶
acados is a software package providing fast and embedded solvers for nonlinear optimal control.
Problems can be conveniently formulated using the CasADi symbolic framework and the high-level acados interfaces.
acados provides a collection of computationally efficient building blocks tailored to optimal control structured problems, most prominently optimal control problems (OCP) and moving horizon estimation (MHE) problems.
Among others, acados implements:
modules for the integration of ordinary differential equations (ODE) and differential-algebraic equations (DAE),
interfaces to state-of-the-art QP solvers like
HPIPM,qpOASES,DAQPandOSQP(partial) condensing routines, provided by
HPIPMnonlinear programming solvers for optimal control structured problems
real-time algorithms, such as the real-time iteration (RTI) and advanced-step real-time iteration (AS-RTI) algorithms
The back-end of acados uses the high-performance linear algebra package BLASFEO, in order to boost computational efficiency for small to medium scale matrices typical of embedded optimization applications.
MATLAB, Octave and Python interfaces can be used to conveniently describe optimal control problems and generate self-contained C code that can be readily deployed on embedded platforms.
Problem Formulation¶
Since acados mainly aims on providing SQP type methods for optimal control, it naturally needs optimal control structured nonlinear programming formulations (OCP-NLP) and quadratic programming (QP) formulations to tackle the subproblems within SQP.
Optimal control structured NLP (OCP-NLP): The problem formulation targeted by
acadosOCP solver is stated here.QP formulations (dense and OCP structured):
acadosrelies onHPIPMfor reformulating QP problems via (partial) condensing and expansion routines. We thus use the flexible QP formulations fromHPIPMfor optimal control structured quadratic programming formulation (OCP-QP) and the dense QP formulation. Both problem formulations are documented in theHPIPM guide.
Documentation page overview¶
Documentation latest build: May 07, 2024