This repository contains the code to our L4DC 2025 paper Diffusion Predictive Control with Constraints. We build upon the temporal U-Net implementation from Diffuser and use the Avoiding environment from D3IL.

Installation

Clone the repo and run:

conda create -n dpcc python=3.10
pip install torch torchvision --index-url https://download.pytorch.org/whl/cu124
pip install -r requirements.txt

You also need to install D3IL for the simulation environment.

Training

To train the diffusion policy, run:

python scripts/train.py

You can also visualize the training data without constraints and the novel test-time constraints:

python scripts/visualize_data_constraints/train.py

Testing

To evaluate DPCC and reproduce the results reported in the paper, run:

python scripts/eval.py
python scripts/load_results.py

Citation

If you find this work useful, please cite our paper:

@InProceedings{romer2025diffusion,
  title = 	 {Diffusion Predictive Control with Constraints},
  author =       {R{\"o}mer, Ralf and Rohr, Alexander von and Schoellig, Angela},
  booktitle = 	 {Proceedings of the 7th Annual Learning for Dynamics \& Control Conference},
  pages = 	 {791--803},
  year = 	 {2025},
  publisher =    {PMLR},
}