buildyn 1.1.0

BuilDyn is an open-source free-to-use PyPI package to help simulating Functional Mockup Units (FMU) for buildings. Its primarily use is probing buildings thermal dynamics, randomizing building parameters, and demand-driven building simulation. As this is an open-source project, we encourage users and researchers to contribute with new ideas and directions to follow for future releases!

Setup

The only hard requirement for this package is python == 3.13.
Aside from that, there are two ways to build this package in your environment:

Installation via Pip

We distribute the newest version of this package over PyPI.

pip install buildyn

Installation from source code

Aside from the pip installation, you can also build the newest version of this package from scratch. First, clone this repository.

git clone https://github.com/felixmkoch/buildyn.git

After that, you can pip install this package into your python environemnt using

pip install .

Strcture

The package has two main classes to consider: FMU and BuilDyn. FMU is a modified and extended wrapper around the FmuSlave from FmPy. BuilDyn enables users full control over variations, walker distributions, and demand-dirven simulation. Multiple functionalities are inspired by BuilDa. For further information about the structure we refer to our paper.

Usage

One general use-case would be the following:

from buildyn import FMU
from buildyn.buildyn import BuilDyn
from buildyn.walker.random.random_walker import RandomWalker
from buildyn.walker.interval_walker import IntervalWalker


fmu_path = "model1.fmu"

start_variables = {
    "weaDat.filNam": "resources/Munich.mos",
    "internalGain.fileName": "resources/NoActivity.txt"
}

fmu = FMU(fmu_path, init_values=start_variables)

observables = ["thermalZone.TAir", "heatingPower"]

rw = RandomWalker(min=10, max=500, is_discrete=True)
iw = IntervalWalker(rw, 900) # 900 is the step size of the walker.

buildyn = BuilDyn(fmu, observables=observables)

buildyn.add_walker_distribution("heatingPower", iw)

print(buildyn.sample_one())

To explore all features of BuilDyn, we refer to our notebooks; each of them thematizes one key part.

Citation

If you are using BuilDyn, consider citing the corresponding paper.

@inproceedings{
  koch26buildyn,
  title={BuilDyn: Excitation-Driven Data Generation for Building Thermal Dynamics Modeling and Control},
  author={Koch, Felix and Krug, Thomas and Raisch, Fabian and Schäfer, Benjamin and Tischler, Benjamin},
  booktitle={Proceedings of the 17th ACM International Conference on Future and Sustainable Energy Systems}
  year={2026}
}