Transfer Learning for Neural Parameter Estimation applied to Building RC Models

TL;DR

This paper presents a transfer learning framework for neural parameter estimation in dynamical systems, specifically applied to building thermal models, achieving significant accuracy improvements with limited training data.

Contribution

It introduces a transfer-learning-based neural estimation method that enhances accuracy and reduces data requirements for building RC models, surpassing traditional approaches.

Findings

Achieved 18.6-24.0% performance improvement with 12 days of data.

Up to 49.4% improvement with 72 days of data.

Outperformed genetic algorithms and from-scratch neural networks.

Abstract

Parameter estimation for dynamical systems remains challenging due to non-convexity and sensitivity to initial parameter guesses. Recent deep learning approaches enable accurate and fast parameter estimation but do not exploit transferable knowledge across systems. To address this, we introduce a transfer-learning-based neural parameter estimation framework based on a pretraining-fine-tuning paradigm. This approach improves accuracy and eliminates the need for an initial parameter guess. We apply this framework to building RC thermal models, evaluating it against a Genetic Algorithm and a from-scratch neural baseline across eight simulated buildings, one real-world building, two RC model configurations, and four training data lengths. Results demonstrate an 18.6-24.0% performance improvement with only 12 days of training data and up to 49.4% with 72 days. Beyond buildings, the proposed method represents a new paradigm for parameter estimation in dynamical systems.