EAGER: Connecting Smart Communities with Intelligent Transportation Systems - Energy Management of Predictive Occupancy Information

建筑和交通是社会中两个相互作用的基本要素。各种交通工具通过现代交通网络在不同的建筑物/社区之间运送居住者。居住者是建筑和社区运作的核心，特别是在能源使用方面。建筑物占用率的预测信息对于提高建筑物能源管理的效率非常有益。当乘客被运送到建筑物/社区时，可以通过智能交通系统（ITS）预测乘客的到达信息，通过车辆与基础设施的通信和数据分析实现。预计这将大大有利于建筑物和社区能源管理的优化。对于乘客携带的插电式电动汽车（pev），这种预测到达信息还可以帮助优化充电管理和车辆到电网的运行。这项早期概念探索性研究补助金（EAGER）支持将智能交通系统与智慧社区的能源管理相结合的探索性研究，即利用智能交通系统预测的入住率信息加强建筑/社区的能源管理。本研究的结果将解决与智能交通与智慧社区融合相关的重大基础问题，有利于技术的发展。该项目的研究结果将通过一个专门的网站向更广泛的社区传播，该网站将展示基于模拟结果的动画和视频剪辑。此外，计划为政府、K-16教师和学生以及一般公众举办研讨会和讲习班，以说明和传播研究结果。本文采用两种社区层面的能源管理方案来展示社区-智能交通系统集成带来的潜在效益：1)需求响应下的社区/区域供冷系统的预测能源管理，通过对即将到来的建筑物住户到达时间的汇总随机估计来增强；2)随机估计即将到来的电动汽车到达时间和到达电池荷电状态（SOC），增强停车场电动汽车的分散充电管理。通过ITS集成实现对进场乘员/车辆到达信息的移动视界预测，对到达时间进行随机预测，同时对PEV到达soc进行随机预测。通过对随机到达时间估计、人员负荷关系动态数据驱动建模和天气信息的聚合，将随机模型预测控制（MPC）策略应用于需求响应运行下的带有中央冷水厂的社区建筑区域供冷。为了了解乘员到达信息的随机估计如何影响随机MPC的最终性能和能量管理问题的随机优化，将讨论不确定性传播的重要问题。

Building and transportation are two essential and interactive elements for the society. Various means of transportation convey occupants among different buildings/communities via modern transportation networks. Occupants are central to operation of buildings and communities, especially in terms of energy use. Predictive information on building occupancy can be dramatically beneficial for improving the efficiency of building energy management. As occupants are transported to buildings/communities, the occupant arrival information can be predicted to through Intelligent Transportation Systems (ITS), enabled by vehicle-to-infrastructure communication and data analytics. This is expected to significantly benefit the optimization of energy management for buildings and communities. For occupant-carrying plug-in electric vehicles (PEVs), such predictive arrival information can also help optimize the charging management and vehicle-to-grid operation. This EArly-concept Grant for Exploratory Research (EAGER) award supports an exploratory research on integrating the ITS to energy management of Smart Communities, i.e. enhancing the building/community energy management with ITS predicted occupancy information. The results of this research will address major fundamental issues relevant to the integration of ITS and Smart Community, which will benefit the technology development. The findings of the project will be disseminated to a broader community via a dedicated website showing animations and video clips based on simulation results. In addition, seminars and workshops for government, K-16 faculty and students, and general public are planned to illustrate and disseminate the results of the research.Two scenarios of community level energy management are used to demonstrate the potential benefits brought by such community-ITS integration: 1) predictive energy management of community/district cooling system under demand response, enhanced with aggregated stochastic estimation of arrival time of upcoming building occupants; 2) decentralized charging management of parking-lot PEV enhanced with stochastic estimation of arrival time and arrival battery state-of-charge (SOC) of upcoming PEVs. With the moving-horizon prediction of in-coming occupant/vehicle arrival information enabled by ITS integration, stochastic prediction of arrival time will be performed, as well as the PEV arrival SOCs. Through aggregation of the stochastic arrival time estimation, dynamic data-driven modeling of occupant-to-load relations, and weather information, a stochastic model predictive control (MPC) strategy is applied for district cooling of community buildings with a central chilled-water plant, under demand response operation. An important issue of uncertainty propagation will be addressed in order to gain the understanding on how the stochastic estimate of occupant arrival information would affect the ultimate performance of the stochastic MPC and stochastic optimization for the energy management problems.