An energy signal tool for decision support in building energy systems
An energy signal tool for decision support in building energy systems
复制标题
DOI：
10.1016/j.apenergy.2014.10.029
复制
发表时间：
2015-01
期刊：
影响因子：
11.2
通讯作者：
G. Henze;G. Pavlak;A. Florita;R. Dodier;A. Hirsch
中科院分区：
文献类型：
--
作者：
G. Henze;G. Pavlak;A. Florita;R. Dodier;A. Hirsch
演示了原型能源信号工具，用于整个建筑和系统级能源使用评估。该工具的目的是提供建筑能源使用情况的摘要，使建筑运营商能够快速区分正常和异常的能源使用。为此，能源使用状态显示为交通灯，这是能源使用的视觉隐喻，与预期（红灯和黄灯）有很大不同，或者与预期（绿灯）大致相同。对于给定的能源最终用途显示哪种光是通过将预期能源使用与实际能源使用进行比较来确定的。由于预期的能源使用必然是不确定的，我们无法确定地选择合适的灯光。相反，能量信号工具通过最小化显示错误光的预期成本来选择光。预期能源使用由概率分布表示。能源使用通过低阶集总参数模型进行建模。通过蒙特卡罗探索模型参数对能源使用的影响来量化能源使用的不确定性。模型参数的分布通过贝叶斯定理随时间更新。模拟研究旨在评估在地下计量水平存在不确定性和故障的情况下整个建筑能源信号的准确性，这可能导致在没有地下计量的情况下无法检测到整个建筑水平的权衡。
A prototype energy signal tool is demonstrated for operational whole-building and system-level energy use evaluation. The purpose of the tool is to give a summary of building energy use which allows a building operator to quickly distinguish normal and abnormal energy use. Toward that end, energy use status is displayed as a traffic light, which is a visual metaphor for energy use which is substantially different from expected (red and yellow lights) or more or less the same as expected (green light). Which light to display for a given energy end-use is determined by comparing expected energy use to actual energy use. As expected energy use is necessarily uncertain, we cannot choose the appropriate light with certainty. Instead the energy signal tool chooses the light by minimizing the expected cost of displaying the wrong light. The expected energy use is represented by a probability distribution. Energy use is modeled by a low-order lumped parameter model. Uncertainty in energy use is quantified by a Monte Carlo exploration of the influence of model parameters on energy use. Distributions over model parameters are updated over time via Bayes’ theorem. The simulation study is devised to assess whole building energy signal accuracy in the presence of uncertainty and faults at the submetered level, which may lead to tradeoffs at the whole building level not detectable without submetering.