The prediction of building heating and ventilation energy consumption base on Adaboost-bp algorithm. Issue 3 (March 2020)

• Record Type: Journal Article
• Authors: Sun, Lunan; Wei, Qi; He, Lipeng; Yin, Zihe
• Is Part Of: IOP conference series. Volume 782:Issue 3(2020)
• Publication Date: 2020-03
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/782/3/032008
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: Particularly In the nowadays, under the environment of increasing severe weather, buildings become consumers of energy resources that cannot be ignored, the hvac is one of the most important energy consuming equipment in the building, it has great practical significance and practical guidance for energy consumption prediction and optimization to reduce overall energy consumption and cost. The Adaboost-BP model based on integrated learning algorithm can not only improve the prediction accuracy of BP neural network algorithm model, at the same time, the defects of BP neural network algorithm such as falling into local minimum and slow convergence speed can be corrected. Moreover, the integrated learning algorithm has low requirements for weak classifiers and almost no need to adjust its parameters, so it has a wide range of use and good robustness. The building cannot be ignored as energy resource consumers, and hvac, as one of the main energy consumption equipment in buildings, prediction and energy consumption in the energy saving optimization to reduce the overall energy consumption, reduce costs. The Adaboost-BP model based on integrated learning algorithm can not only improve the prediction accuracy of BP neural network algorithm, but also correct the defects of BP neural network algorithm such as falling into local minimum value and slow convergence speed. Moreover, the integrated learning algorithm has very low requirements for weak classifiers and almost no need to adjust its parameters, so it has a wide range of use and good robustness .In conclusion, the energy consumption forecasting and optimization scheduling based on data-driven have good effect to optimize the energy consumption structure of office buildings, save energy resources, reduce greenhouse gas emissions, and reduce the impact on the power grid caused by the increase of demand from users during the peak period of electricity consumption, also provides a design idea for distributed energy network design.