密集区域民居建筑整体温度调节能耗预测
|
摘要
针对当前能耗预测时,普遍存在预测时间较长、拟合优度较低、成本开销较大等问题,提出基于人工鱼群算法的能耗预测方法.引入时间序列方法寻找出附近2个变量之间温度能耗密切程度,获取建筑温度能耗相关函数,得出建筑温度能耗季节关联性,采用人工鱼群算法对神经网络阈值进行优化,构建预测模型,以此完成能耗预测.实验结果表明,所提出方法预测时间较短,成本消耗较低,拟合优度较高.
The current temperature regulation energy consumption forecast of residential buildings in dense areas has long prediction time, low goodness of fit, and high cost. An energy consumption prediction method based on artificial fish swarm algorithm is proposed. The time series method is introduced to find the temperature energy consumption between two adjacent variables. The correlation function of building temperature and energy consumption and the seasonal correlation of building temperature and energy consumption were obtained. The artificial fish swarm algorithm was used to optimize the threshold of BP neural network, and the overall temperature energy consumption prediction model of the building was constructed. The temperature regulation energy consumption prediction of dense residential buildings was completed. The experimental results show that the prediction time required in the consumption prediction is shorter, the cost is lower, and the goodness of fit is higher than the current temperature regulation energy consumption forecast of residential buildings.
The current temperature regulation energy consumption forecast of residential buildings in dense areas has long prediction time, low goodness of fit, and high cost. An energy consumption prediction method based on artificial fish swarm algorithm is proposed. The time series method is introduced to find the temperature energy consumption between two adjacent variables. The correlation function of building temperature and energy consumption and the seasonal correlation of building temperature and energy consumption were obtained. The artificial fish swarm algorithm was used to optimize the threshold of BP neural network, and the overall temperature energy consumption prediction model of the building was constructed. The temperature regulation energy consumption prediction of dense residential buildings was completed. The experimental results show that the prediction time required in the consumption prediction is shorter, the cost is lower, and the goodness of fit is higher than the current temperature regulation energy consumption forecast of residential buildings.
引文
[1]耿宏,胡小娜,陈静杰.基于小波-ARIMA的航空运输企业能耗预测模型[J].机床与液压,2018,46(6):13-17,42.
[2]赵晓华,姚莹,伍毅平,等.基于主成分分析与BP神经元网络的驾驶能耗组合预测模型研究[J].交通运输系统工程与信息,2016,16(5):185-191,204.
[3]马晓雯,刘雄伟,刘刚,等.深圳市建筑能耗宏观影响因素分析及发展趋势情景预测[J].暖通空调,2017,47(6):15-20,132.
[4]王仁群,彭力.数据中心网络拓扑感知型能耗优化算法[J].计算机工程与应用,2017,53(17):117-122.
[5]樊丽军.基于多元线性回归模型的建筑能耗预测与建筑节能分析[J].湘潭大学自然科学学报,2016,38(1):123-126.
[6]周峰,张立茂,秦文威,等.基于SVM的大型公共建筑能耗预测模型与异常诊断[J].土木工程与管理学报,2017,34(6):80-86.
[7]侯立强,杨柳,李红莲,等.气象参数对成都地区办公建筑能耗的影响及预测[J].土木建筑与环境工程,2017,39(4):56-62.
[8]程文荣,姚天祥.南京市化学原料及化学制品制造业能源消耗预测[J].工业安全与环保,2016,42(1):89-92,99.
[9]蔡宙燊,张昕,张宇涛.中国便利店的照明能耗预测模型与主观评价[J].土木建筑与环境工程,2017,39(4):48-55.
[10]胡慧,王桂芝,来鹏.基于PLSIM模型的住房建筑物能耗分析[J].数理统计与管理,2016,35(5):770-777.
[2]赵晓华,姚莹,伍毅平,等.基于主成分分析与BP神经元网络的驾驶能耗组合预测模型研究[J].交通运输系统工程与信息,2016,16(5):185-191,204.
[3]马晓雯,刘雄伟,刘刚,等.深圳市建筑能耗宏观影响因素分析及发展趋势情景预测[J].暖通空调,2017,47(6):15-20,132.
[4]王仁群,彭力.数据中心网络拓扑感知型能耗优化算法[J].计算机工程与应用,2017,53(17):117-122.
[5]樊丽军.基于多元线性回归模型的建筑能耗预测与建筑节能分析[J].湘潭大学自然科学学报,2016,38(1):123-126.
[6]周峰,张立茂,秦文威,等.基于SVM的大型公共建筑能耗预测模型与异常诊断[J].土木工程与管理学报,2017,34(6):80-86.
[7]侯立强,杨柳,李红莲,等.气象参数对成都地区办公建筑能耗的影响及预测[J].土木建筑与环境工程,2017,39(4):56-62.
[8]程文荣,姚天祥.南京市化学原料及化学制品制造业能源消耗预测[J].工业安全与环保,2016,42(1):89-92,99.
[9]蔡宙燊,张昕,张宇涛.中国便利店的照明能耗预测模型与主观评价[J].土木建筑与环境工程,2017,39(4):48-55.
[10]胡慧,王桂芝,来鹏.基于PLSIM模型的住房建筑物能耗分析[J].数理统计与管理,2016,35(5):770-777.