SBIR Phase I: A Data-driven Demand Response Recommendation System

这个小企业创新研究（SBIR）第一阶段项目的更广泛的影响/商业潜力将是为建筑物和电网运营商提供一个数据智能层，并帮助消除对优化电力需求响应实施的任何猜测。更好的需求响应意味着，在高电力需求时期，效率较低、往往成本较高的发电形式不需要上线。需求响应优化减少了输配电系统的压力，使它们不太可能发生故障。良好实施的需求响应计划可促进竞争性和高效率的批发电力市场运作，并有助于控制电价。这项小企业创新研究（SBIR）第一阶段项目旨在开发、部署和评估能源系统面向控制的预测建模的创新算法和软件，并具体应用于电力需求响应。提议的工作将解决四个重要的技术障碍：(a)建模的复杂性和异质性，(b)基于规则的需求响应的局限性，(c)控制的复杂性和可扩展性，以及(d)建模和控制的可解释性。拟议研究的目标是解决每一个技术障碍，并在此过程中开发(i)面向需求响应和大规模能源系统的数据驱动控制建模工具，以及（ii）一种方法，以保持我们的数据驱动控制导向模型的可解释性，为设施经理提供基于现实世界数据的能源分析。建议的计划将包括在实际楼宇进行试验部署，以评估需求回应软件的表现。

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project will be to provide a data intelligence layer for buildings and electric grid operators, and help remove any guesswork from the implementation of optimized electricity demand response. Better demand response means that less efficient, and often more expensive, forms of electricity generation do not need to come online during times of high electricity demand. Demand response optimization reduces the stress on transmission and distribution systems, making them less likely to fail. A well implemented demand response program leads to competitive and efficient wholesale electricity market operations, and helps in keeping electricity prices in check. This Small Business Innovation Research (SBIR) Phase I project aims to develop, deploy, and evaluate innovative algorithms and software for control-oriented predictive modeling for energy systems, with specific application to electricity demand response. There are four significant technical hurdles that will be addressed by the proposed effort: (a) modeling complexity and heterogeneity, (b) limitations of rule-based demand response, (c) control complexity and scalability, and (d) interpretability of modeling and control. The goal of the proposed research is to address each of the technological hurdles, and in doing so develop (i) data-driven control oriented modeling tools for demand response and for large scale energy systems, and (ii) a methodology to preserve interpretability of our data-driven control-oriented models for providing energy analytics to the facilities manager based on real-world data. The proposed project will include pilot deployments on real buildings to evaluate the performance of the demand response software.