A ROBUST ENSEMBLE LEARNING FRAMEWORK FOR FORECASTING AIR CONDITIONING ENERGY CONSUMPTION IN EDUCATIONAL BUILDINGS
TL;DR
Results confirm the effectiveness, reliability, and practical application potential of the model in the problem of building electricity load forecasting, and opens up prospects for developing advanced forecasting and assessment models to serve energy management and ensure energy security in the future.
Abstract
In the context of the growing development of large-scale building energy monitoring platforms, accurate forecasting of electricity demand plays a crucial role in optimizing system design, improving operational efficiency, energy management, and supporting the integration of renewable energy into the grid. In buildings, air conditioning systems often account for a large proportion of the total electrical load; therefore, effective monitoring and forecasting of their consumption is particularly important. This study proposes a new computational framework to forecast the daily electricity consumption trends of school buildings by integrating time-dependent characteristics with modern combined machine learning algorithms. Experimental results show that the proposed method outperforms existing methods, achieving MAPE = 35.2%. These results confirm the effectiveness, reliability, and practical application potential of the model in the problem of building electricity load forecasting. At the same time, the research also opens up prospects for developing advanced forecasting and assessment models to serve energy management and ensure energy security in the future.