Transfer Learning for Neural Parameter Estimation applied to Building RC Models
This addresses the problem of efficient and accurate parameter estimation for building thermal models and potentially other dynamical systems, representing a new paradigm rather than an incremental improvement.
The paper tackles the challenge of parameter estimation in dynamical systems by introducing a transfer-learning-based neural framework, which improves accuracy by 18.6-24.0% with 12 days of training data and up to 49.4% with 72 days, while eliminating the need for initial parameter guesses.
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.