To main content

LandSkape - Hybrid Physical-Based Deep Learning for Fast and Reliable Wind Flow Estimation

Helping to ensure the overall well-being, safety, and comfort in pedestrian zones in building assessment is very important. Current technology used for wind flow calculation results in accurate assessment of wind comfort but has the drawback of being extremely time consuming and then not suitable for fast iteraction for end users like architects or urban planner. We believe that current AI tecniques can help in defining surrogate models able to quickly estimate the output of those simulations and then useful as core methods tools for smart building assessment frameworks.

Contact person

Ill.: SINTEF

With the complexity of modern urban areas, the pedestrian wind environment analysis becomes a critical factor in urban and building planning design, helping to ensure the overall well-being, safety, and comfort in pedestrian zones. The usage of fluid flow simulation enables architects and engineers to predict and optimize the performance of buildings in the early stage of the design process. In general those traditional Computaional Fluid Dynamics (CFD) methods produce high-accuracy results, but they are computationally expensive and does not work well in the design process of new prototypes in a given domain. To obtain results, it often takes several hours or days, depending on the prototype’s complexity.

The main objective of the project is to explore the potential of Modern ML (Machine Learning) techniques as approximation of those simulations. Deep Neural Network-based architectures represent a fast and alternative solution for efficiently approximating mapping function in high dimensional spaces. The main advantage of using DNN-based architecture as a surrogate model, will be:

  • having a model with a high degree of generalization
  • quick training time (different magnitude in comparison to CFD simulation)
  • quick inference with the advantage of interaction during the design process

We aim to explore deep learning with the objective of creating an interactive tool for testing new designs, even when they are getting computationally hard for physical solvers. In particular, we plan to go in a similar direction and define DL-based architectures (eventually physical-informed) that can generate wind flows for arbitrarily shaped buildings in scenario of different complexity (city maps) with the motivation of building a surrogate model that can be used in an interactive tool for smart building assessments.

In line with the green shift, the project will impact urban and building planning design, helping to ensure the overall well-being, safety, and comfort in pedestrian zones. Better site assessment and insight in wind efficiency, with the corresponding cost reduction for urban wind park developments, can also positively affect societal acceptance and will allow the user to assess the sites and investigate minimal environmental impact by positioning of the structures. The project contributes to UN SDGs #7, #9 and #13.

Key facts

Project duration

2021 - 2023

Funding

The Research Council of Norway

Partners

Nablaflow AS (Project Owner)
Department of Mechanical and Structural Engineering and Materials Science (UiS)
Department of Architecture and Technology (NTNU)
Stavanger Kommune
Nordic Edge

Project type

IPN