🧪 Test?View on arXiv
Comparative review of hybrid forecasting models for short-term prediction of building thermal load
forecastingthermal loadhybrid models
2610.06881
Builder Relevance
1h ago80%
Abstract
This paper reviews various hybrid models for short-term forecasting of building thermal demand, assessing their performance using historical data.
Reality Card
Core Claim
The EMD-LSTM-Markov model achieves the highest accuracy in predicting day-ahead heating and domestic hot water demands among 13 hybrid methods evaluated.
Method / Result
The EMD-LSTM-Markov model predicts heating and DHW demands with the highest accuracy, though it underestimates high power swells and spikes.
Limitations
High power swells and spikes are underestimated by the best-performing model, which may affect practical applications.
Paper to code
Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.
No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.