期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
基于混合深度学习模型的企业中央空调蒸汽预测
1
作者 王可 卢焕达 +1 位作者 郑军红 何利力 《南京师范大学学报(工程技术版)》 CAS 2022年第3期53-62,共10页
为解决企业生产车间多空调能耗与生产任务、气候环境匹配的精准供能问题,实现多台大型中央空调机组蒸汽消耗预测,提出一种基于GRU和3DConv-PredRNN++的混合深度学习预测模型.针对多台空调机组动态联动关系,使用三维卷积和PredRNN++方法... 为解决企业生产车间多空调能耗与生产任务、气候环境匹配的精准供能问题,实现多台大型中央空调机组蒸汽消耗预测,提出一种基于GRU和3DConv-PredRNN++的混合深度学习预测模型.针对多台空调机组动态联动关系,使用三维卷积和PredRNN++方法提取机组间蒸汽损耗关系作为空间因素特征参与模型预测;为捕捉蒸汽消耗量序列的总体趋势和局部变化,数据集采用平滑过程模式、趋势性模式和周期性模式作为模型输入;为提高模型预测性能,基于门控循环单元(GRU)耦合外部因素特征并捕捉时间因素特征;最后通过参数矩阵融合方式来构建模型.通过与多种预测模型的对比实验,证明混合深度学习模型预测精度的优越性和空间因素特征参与模型预测的必要性.与现有模型相比,所提模型平均能耗折标(ASEC)降低了60.09%. 展开更多
关键词 空调蒸汽消耗 三维卷积 PredRNN++ 门控循环单元 空间特征
下载PDF
你知道这些耗油数字吗
2
作者 苏杭 《汽车运用》 2004年第4期26-26,共1页
关键词 燃油消耗 怠速运转耗油 化油器 量孔调整 空调消耗 消耗 电力系统消耗
下载PDF
Load Management Scheme Using Air Conditioning Electric Power Consumption and Photovoltaic Power System Generation
3
作者 Zvonimir Klaic Damir Sljivac Kresimir Fekete Zorislav Kraus 《Journal of Energy and Power Engineering》 2014年第11期1926-1932,共7页
In the last few decades, in the world and also in the European Union, considerable resources had been invested in the rapid development of renewable energy sources and distributed generation in general. At the same ti... In the last few decades, in the world and also in the European Union, considerable resources had been invested in the rapid development of renewable energy sources and distributed generation in general. At the same time, power consumption is continuously increasing, and consumers are becoming more complex, which ultimately requires new investments in the distribution network. Concept of smart grids is generally accepted as a possible solution. Smart grid is a concept with many elements, where monitoring and control of every element in the chain of production, transmission, distribution and final consumption enable much more efficient delivery and use of electricity. One of the elements of smart grid efficiency is the ability of real-time demand-supply balancing. This balancing is carried out by monitoring of consumption and redistribution of electricity among individual end users, according to their needs. The aim of this paper is creating algorithm for real-time load management using power measurements. Algorithm for real-time load management at the ETFOS (Faculty of Electrical Engineering in Osijek), Croatia is created based on measurements of photovoltaic power plant production, the power consumption of air conditioning system and the faculty building total electricity consumption. Expected result of real-time re-dispatching of air conditioners consumption, depending on the level of electricity production in photovoltaic power plant is decreasing peak demand of the faculty. 展开更多
关键词 Photovoltaic power plant smart grid load management air conditioning system power measurements.
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部