Improving Passenger Thermal Comfort using AI

Urvi Mehta
Urvi Mehta New Altair Community Member
edited May 18 in Other Discussion & Knowledge

Introduction

In the modern technological landscape, digital twin technology is revolutionizing the way we interact with physical objects and systems. A digital twin is a computerized model of an object or system that covers its entire lifecycle, utilizing real-time data for constant updates. This article demonstrates the integration of Computational Fluid Dynamics (CFD) results into a 1D Twin Activate HVAC model using Reduced Order Models (ROMs) created with romAI, aimed at predicting temperature, relative humidity, and power consumption in a car cabin.

Objective: Integrate a 3D CFD model of an automotive cabin into a 1D HVAC system model using romAI to combine detailed simulation data with efficient modeling techniques.

Key Question: Can romAI predict the temperature curve in the cabin?

Methodology

image

  • CFD Simulations: Using AcuSolve, perform CFD simulations with velocity and temperature values as inputs and output temperature values at required points.
  • ROM Creation: Use the temperature data to create ROMs with romAI.
  • Model Testing: Test the ROMs for new cases and compare results with CFD-generated plots.
  • Deployment: Deploy the ROMs in the 1D systems model.

CFD Analysis

image

Creation of Reduced Order Models in romAI

image

Validating/Testing ROMs

image

  • Testing Procedure: Run unseen cases with varied velocity and temperature, compare results with CFD simulations.
  • Outcome: Significant reduction in prediction time from 12 hours to less than a minute with romAI.

Incorporating ROMs in the HVAC Model

  • Integrate ROMs into Twin Activate "HVAC“ model and calculate overall cabin temperature by averaging temperatures of the four nodes.
  • Plot data for average cabin temperature, environment temperature, and reference temperature as seen in the video attachment.

 

Tagged:

Categories