摘要
针对太阳能-土壤源热泵复合系统在水箱直接供暖模式下运行,存在热泵机组频繁启停以及供暖负荷分配不合理的问题,提出一种基于BP神经网络的动态供暖策略。以沈阳某建筑为对象,设计太阳能-土壤源热泵复合系统,通过TRNSYS和Matlab软件进行仿真。通过与常规水箱直接供暖策略对比,该策略能在典型供暖日中将机组的启停次数从9次减少为2次,机组COP从3.90提高至4.02。在供暖负荷较大的供暖季中期,平均每天提高机组COP 2.37%。
Aiming at the problems of frequent shutdown and unreasonable heating load distribution when running direct heating mode using water tank in the solar-ground source heat pump composite system,a dynamic heating strategy based on BP neural network was proposed.Taking a building in Shenyang as an research object,a solar-ground source heat pump composite system was designed,and simulated by using TRNSYS and Matlab software.Compared with the traditional direct heating mode using water tank,this strategy can reduce the on/off number of the unit from 9 to 2,and increase the COP of the unit from 3.90 to 4.02 during a typical heating day.In the middle of heating season,the COP of the unit is increased by 2.37%on average per day.
作者
杨震
陈翔燕
刘诚
许波
吕钟灵
陈振乾
Yang Zhen;Chen Xiangyan;Liu Cheng;Xu Bo;Lyu Zhongling;Chen Zhenqian(School of Energy and Environment,Southeast University,Nanjing 210096,China;Jiangsu Shengshi Mechanical and Electrical Engineering Co.,Ltd.,Lianyungang 222000,China;China Design Group,Nanjing 210014,China)
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2022年第8期224-229,共6页
Acta Energiae Solaris Sinica
基金
华设设计集团股份有限公司开放基金(8503008464)。
关键词
地源热泵
太阳能
神经网络
供暖策略
ground source heat pumps
solar energy
neural networks
heating strategy
