You are required to read and agree to the below before accessing a full-text version of an article in the IDE article repository.

The full-text document you are about to access is subject to national and international copyright laws. In most cases (but not necessarily all) the consequence is that personal use is allowed given that the copyright owner is duly acknowledged and respected. All other use (typically) require an explicit permission (often in writing) by the copyright owner.

For the reports in this repository we specifically note that

  • the use of articles under IEEE copyright is governed by the IEEE copyright policy (available at http://www.ieee.org/web/publications/rights/copyrightpolicy.html)
  • the use of articles under ACM copyright is governed by the ACM copyright policy (available at http://www.acm.org/pubs/copyright_policy/)
  • technical reports and other articles issued by M‰lardalen University is free for personal use. For other use, the explicit consent of the authors is required
  • in other cases, please contact the copyright owner for detailed information

By accepting I agree to acknowledge and respect the rights of the copyright owner of the document I am about to access.

If you are in doubt, feel free to contact webmaster@ide.mdh.se

Learning in modal space: Physics-guided diffusion for multi-scale time-series generation

Authors:

Zafer Yigit, Håkan Forsberg, Masoud Daneshtalab

Research group:

Heterogeneous systems - hardware software co-design

Publication Type:

Conference/Workshop Paper

Venue:

31st IEEE International Conference on Emerging Technologies and Factory Automation

Abstract

Generating realistic multivariate time-series data remains challenging in complex dynamical systems where heterogeneous processes evolve across multiple time scales and are entangled in the original signal space. This study proposes a physics-guided generative framework that combines Dynamic Mode Decomposition with a transformer-based diffusion model to restructure the representation space and separate multi-scale dynamics into coherent components. The diffusion model is trained in this modal space and the generated sequences are reconstructed back to the original domain. Experimental results show that the proposed approach produces smoother temporal behaviour, better preserves spectral characteristics, and more accurately captures temporal and cross-variable dependencies compared with baseline models. The structured modal representation also reduces learning complexity, resulting in an approximately threefold reduction in training time. These results demonstrate that the proposed approach generates more realistic and physically consistent time-series data.

Bibtex

@inproceedings{Yigit7414,
author = {Zafer Yigit and H{\aa}kan Forsberg and Masoud Daneshtalab},
title = {Learning in modal space: Physics-guided diffusion for multi-scale time-series generation},
month = {September},
year = {2026},
booktitle = {31st IEEE International Conference on Emerging Technologies and Factory Automation},
url = {http://www.es.mdu.se/publications/7414-}
}