MPC test set for QP solvers

This repository contains quadratic programs (QPs) arising from model predictive control in robotics, in a format suitable for qpbenchmark. Here is the report produced by this benchmarking tool:

📈 MPC test set results

Installation

The recommended process is to install the benchmark and all solvers in an isolated environment using conda:

conda env create -f environment.yaml
conda activate mpc_qpbenchmark

It is also possible to install the benchmark from PyPI.

Adding HPIPM to the conda environment

HPIPM is not packaged, but instructions to install from source are given in hpipm:

Clone BLASFEO: git clone https://github.com/giaf/blasfeo.git
From the BLASFEO directory, run: make shared_library -j 4
Check again that you are in your conda environment, then run:

cp -f ./lib/libblasfeo.so ${CONDA_PREFIX}/lib/
cp -f ./include/*.h ${CONDA_PREFIX}/include/

Clone HPIPM: git clone https://github.com/giaf/hpipm.git
From the HPIPM directory, run: make shared_library -j4 BLASFEO_PATH=${CONDA_PREFIX}
Check again that you are in your conda environment, then run:

cp -f libhpipm.so ${CONDA_PREFIX}/lib/
cp -f ./include/*.h ${CONDA_PREFIX}/include/

Go to hpipm/interfaces/python/hpipm_python and run pip install .
Try to import the package in Python:

import hpipm_python.common as hpipm

Usage

Run the test set as follows:

python ./mpc_qpbenchmark.py run

The outcome is a standardized report comparing all available solvers against the different benchmark metrics. You can check out and post your own results in the Results forum.

Contributions

The problems in this test set have been contributed by:

Problems	Contributor	Details
`QUADCMPC*`	@paLeziart	Proposed in #1, details in this thesis
`LIPMWALK*`	@stephane-caron	Proposed in #3, details in this paper
`WHLIPBAL*`	@stephane-caron	Proposed in #4, details in this paper

Limitations

Here are some known areas of improvement for this benchmark:

Cold start only: we don't evaluate warm-start performance for now.
CPU thermal throttling: the benchmark currently does not check the status of CPU thermal throttling.
- Adding this feature is a good way to start contributing to the benchmark.

Note that this test set was spun off to benefit from the availability of qpbenchmark and readily-available MPC QPs, but it does not fully reflect the use of QP solvers for MPC in production due, notably, to the cold-start-only limitation.

Citation

If you use qpbenchmark in your works, please cite all its contributors as follows:

@software{qpbenchmark2024,
  title = {{qpbenchmark: Benchmark for quadratic programming solvers available in Python}},
  author = {Caron, Stéphane and Zaki, Akram and Otta, Pavel and Arnström, Daniel and Carpentier, Justin and Yang, Fengyu and Leziart, Pierre-Alexandre},
  url = {https://github.com/qpsolvers/qpbenchmark},
  license = {Apache-2.0},
  version = {2.3.0},
  year = {2024}
}

If you contribute to this repository, don't forget to add yourself to the BibTeX above and to CITATION.cff.

qpsolvers/mpc_qpbenchmark