Distributed buildings energy storage charging load forecasting method considering parallel deep learning model. (19th November 2019)

• Record Type: Journal Article
• Authors: Yang, Shengying; Wu, Jianfeng; Qin, Huibin; Xie, Qiangqiang; Xu, Zhiwang; Hua, Yongzhu
• Is Part Of: Concurrency and computation. Volume 33:Number 12(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-11-19
• Subjects: building energy -- distributed system -- energy storage -- load forecast -- parallel architecture; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.5580
• Languages: English
• ISSNs: 1532-0626

Abstract

Summary: At present, with total building energy consumption accounts for about 21.33% of the energy consumption of social terminals, the building energy consumption has a tendency to continue to grow. According to the geographical location of users and the local weather conditions, we analyze the energy resources available to users and design a multienergy complementary coupling system. We have taken full account of the use of renewable energy and recovery of waste heat resources. Large‐scale increase of electrical equipment, a large number of charging load access to the grid, the power system planning, operation and operation of the electricity market will have a profound impact. The current mode of the supercapacitor and the DC bus determines the mode of operation of the converter. Based on a detailed analysis of each working mode, we design a corresponding control scheme and achieve a smooth transition and switching between modes. Simulation and experiment verify the correctness and effectiveness of the converter and hybrid energy storage control strategy.