Modelling Computational Fluid Dynamics with Swarm Behaviour

Jankovic, Ljubomir

Abstract

The paper looks at replacing the current top-down approach to modelling, predominantly used in dynamic simulation tools, with a nature inspired bottom-up approach based on principles of swarming. Computational fluid dynamics (CFD) is chosen for this research, as one of the most time-consuming processes under the traditional simulation approach. Generally based on Navier-Stokes simultaneous differential equations, CFD requires considerable user preparation time and considerable CPU execution time. The main reason is that the top-down equations represent the system as a whole and generate a large solution space, requiring a solver to find a solution. However, air and building materials do not have cognitive capabilities to solve systems of equations in order to {\textquoteleft}know{\textquoteright} how to transfer heat. Instead, heat transfer occurs through proximity interaction between molecules, leading to self-organised behaviour that is much faster than the behaviour modelled by the top-down systems of equations. The paper investigates how the bottom-up approach using the principles of swarming could improve the speed and interactivity of CFD simulation.

• Presentation Conference Type: Conference Paper (published)
• Conference Name: 4th IBPSA-England Conference on Building Simulation and Optimization 2018
• Start Date: Sep 11, 2018
• End Date: Sep 12, 2018
• Publication Date: Sep 11, 2018
• Deposit Date: Oct 11, 2024
• Pages: 112-118
• Book Title: Proceedings of the 4th IBPSA-England Conference on Building Simulation and Optimization
• Publisher URL: https://publications.ibpsa.org/conference/?id=bso2018