Building thermal load prediction through shallow machine learning and deep learning

Abstract

Building thermal load prediction informs the optimization of cooling plant and thermal energy storage. Physics-based prediction models of building thermal load are constrained by the model and input complexity. In this study, we developed 12 data-driven models (7 shallow learning, 2 deep learning, and 3 heuristic methods) to predict building thermal load and compared shallow machine learning and deep learning. The 12 prediction models were compared with the measured cooling demand. It was found XGBoost (Extreme Gradient Boost) and LSTM (Long Short Term Memory) provided the most accurate load prediction in the shallow and deep learning category, and both outperformed the best baseline model, which uses the previous day's data for prediction. Then, we discussed how the prediction horizon and input uncertainty would influence the load prediction accuracy. Major conclusions are twofold: first, LSTM performs well in shortterm prediction (1h ahead) but not in long term prediction (24h ahead), because the sequential information becomes less relevant and accordingly not so useful when the prediction horizon is long. Second, the presence of weather forecast uncertainty deteriorates XGBoost's accuracy and favors LSTM, because the sequential information makes the model more robust to input uncertainty. Training the model with the uncertain rather than accurate weather data could enhance the model's robustness. Our findings have two implications for practice. First, LSTM is recommended for short-term load prediction given that weather forecast uncertainty is unavoidable. Second, XGBoost is recommended for long term prediction, and the model should be trained with the presence of input uncertainty.

Classification of retrofit measures for residential buildings according to the global cost

“…The predictive performance of various ML methods is verified and compared in refs. [13,16,17,18]. These methods can be further improved by hybridizing and implementing optimization [14,19].…”

“…Short-term and time-series prediction of energy demand [26] is performed in refs. [18,27,28,29]. Reference [30] predicts the indoor air temperature and humidity in an industrial building.…”

“…Reference [30] predicts the indoor air temperature and humidity in an industrial building. Dominant methods are recurrent ANN and extreme GB [18]. Supervised ML methods and short-term predictions are widelyused in data-driven control applications, e.g.…”

Global sensitivity analysis of two‐stage thermoelectric refrigeration system based on response variance

“…It has been widely used in electronic equipment cooling, waste heat recovery and small refrigeration devices 8‐10 . In recent years, building energy conservation and indoor comfort has been received extensive attention 11,12 . Using renewable energy to control indoor environment and comfort in buildings can achieve building energy conservation.…”

Study on dynamic thermal characteristics of thermoelectric radiant cooling panel system through a hybrid method

“…At the same time, we noticed that machine learning method such as artificial neural network has been showing its power in solving some difficult problems that conventional is incapable or ineffective. The machine learning methods are found to be highly powerful for building energy prediction [ 32 , 33 ], and some researches starts using ANN for thermoelectric based system evaluation [ 34 ] or thermoelectric generator [ 35 ]. However, very few has shown its application for radiant cooling systems.…”