Component-Based Machine Learning for Indoor Flow and Temperature Fields Prediction Latent Feature Aggregation and Flow Interaction

2025-07-25

AI Summary

Traditional computational fluid dynamics (CFD) simulations incur prohibitive computational costs, hindering real-time optimization and rapid design iteration in building environmental engineering. Method: This paper proposes a modular machine learning surrogate model for fast and accurate prediction of indoor airflow and temperature fields. The architecture synergistically integrates a convolutional autoencoder (CAE), multilayer perceptron (MLP), and convolutional neural network (CNN), augmented with latent feature aggregation and flow-field interaction mechanisms to enable scalable modeling—from single- to multi-inlet configurations. Residual connections and feature compression are incorporated to enhance physical consistency and generalization. Results: Evaluated on a 2D room with dual inlets, the model achieves over 100× speedup versus CFD while maintaining high accuracy and robustness on both training and test sets. It provides an efficient, reliable alternative for intelligent building environment design and optimization.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Accurate and efficient prediction of indoor airflow and temperature distributions is essential for building energy optimization and occupant comfort control. However, traditional CFD simulations are computationally intensive, limiting their integration into real-time or design-iterative workflows. This study proposes a component-based machine learning (CBML) surrogate modeling approach to replace conventional CFD simulation for fast prediction of indoor velocity and temperature fields. The model consists of three neural networks: a convolutional autoencoder with residual connections (CAER) to extract and compress flow features, a multilayer perceptron (MLP) to map inlet velocities to latent representations, and a convolutional neural network (CNN) as an aggregator to combine single-inlet features into dual-inlet scenarios. A two-dimensional room with varying left and right air inlet velocities is used as a benchmark case, with CFD simulations providing training and testing data. Results show that the CBML model accurately and fast predicts two-component aggregated velocity and temperature fields across both training and testing datasets.

Problem

• Predict indoor airflow and temperature efficiently
• Replace slow CFD simulations with machine learning
• Aggregate latent features for multi-inlet scenarios

Innovation

• Component-based ML replaces CFD for fast prediction
• Uses CAER, MLP, CNN for feature extraction and aggregation
• Accurate dual-inlet flow and temperature field prediction