Machine learning augmented simulation of urban wind flows

Engineers that design and enhance buildings widely rely on computer simulations to understand wind flows. Wind flows around buildings can affect human safety, comfort, structural safety, and the surrounding environment.
Therefore, accurate and fast simulation of urban wind flows is critical to ensure safe, comfortable, and sustainable designs. If architectural engineers have access to a fast and accurate tool for simulating urban window flows, they can rapidly iterate and design the most energy efficient and sustainable buildings. Though robust, the current methodology for fast urban wind flow simulations often is inaccurate, meaning that slow and expensive simulations are required, slowing down the design process. The goal of this project is to use machine learning to improve current fast methodologies, thereby giving architectural engineers enhanced capabilities to design the safer and more sustainable buildings of the future.

Faculty Supervisor:

Fue-Sang Lien

Student:

Partner:

Rowan Williams Davies & Irwin Inc

Discipline:

Engineering

Sector:

Construction and infrastructure; Professional, scientific and technical services

University:

University of Waterloo

Program: