智能电网环境下家庭能源管理系统调度算法研究

CORC > 沈阳自动化研究所 > 中国科学院沈阳自动化研究所 > 工业控制网络与系统研究室

题名	智能电网环境下家庭能源管理系统调度算法研究
作者	张延宇
学位类别	博士
答辩日期	2015-12-01
授予单位	中国科学院沈阳自动化研究所
授予地点	中国科学院沈阳自动化研究所
导师	曾鹏
关键词	智能电网需求响应家庭能源管理系统调度算法粒子群优化
其他题名	Research on scheduling algorithm for home energy management system in smart grid
学位专业	机械电子工程
中文摘要	本文主要研究内容和取得的成果如下：（1）研究了智能电网环境下居民侧空调系统多目标优化调度算法。首先提出了智能电网环境下空调系统运行的新模式。然后，提出了一种度量用户温度舒适度的指标，并建立了一种同时考虑用户舒适度和用电费用的空调系统多目标优化调度模型。在建模过程中，利用情景分析法处理了室外温度预测误差带来的不确定性，采用改进粒子群算法对所建模型进行求解。最后，通过仿真实验对算法进行了验证。仿真结果表明所提算法能够有效将空调系统的部分负荷从高电价时段转移到低电价时段，减少了用户用电费用和降低电力系统峰值负荷，提高电力系统的稳定性；多目标算法为用户在舒适度和经济性之间的灵活控制提供了一种有效手段。（2）研究了智能电网环境下居民用户作为电能产消者的家庭微电网调度算法。首先，提出了智能电网环境下家庭微电网构成框架、能量交换模型，该框架综合考虑了电能生产、存储、消费、向电网卖电能力对用户的影响。然后，按照设备运行特性和用户偏好，分别为空调系统、电热水器、电动汽车、洗衣机、干衣机和洗碗机等家庭环境内的主要可调度用电设备建立了舒适度指标；基于这些指标建立了能够同时最优化用户舒适度和用电费用的家庭微电网多目标优化调度模型；最后提出了家庭微电网多目标优化调度算法，并进行了仿真实验验证。仿真结果表明本文算法能有效地实现负荷的削峰填谷，同时兼顾用户的舒适度和用电费用，有利于提高可再生能源的利用率，支持在居民侧实现需求响应。（3）研究了智能电网环境下在家庭微电网的框架下实现电动汽车V2H功能的调度算法。首先，提出了含有V2H功能的家庭能源管理系统的组成结构，分析了电动汽车不同的接入状态和荷电状态下，家庭能源管理系统各组成部件间的能量分配关系；然后，建立了含有V2H功能的家庭能源管理系统最优调度模型；最后提出了含有V2H功能的家庭能源管理系统优化调度算法，并进行了仿真验证。仿真结果表明通过利用电动汽车的V2H功能，在高电价时段将电动汽车储存的电能供给家庭环境内的其他负载使用，可有效降低用户的用电费用。（4）以电动汽车充电过程为例，研究了智能电网环境下多用户协同控制消除RTP电价机制下多用户独自优化调度时产生的负荷反弹效应的调度算法。首先提出了智能电网环境下多电动汽车的协同充电方式；然后，建立了各个电动汽车的最优充电模型；最后，提出了一种基于注水算法和基于充电功率按比例分配充电容量的多电动汽车协同充电调度算法，并进行了仿真验证。仿真结果表明该算法能够有效避免变压器负载超限；与现行的超载时才向用户发送DR（Demand Response, DR）信号的方法比，本文算法控制下更多用户能够在调度结束时电动汽车SOC达到指定值，并且充电费用低于现行方法下的充电费用。（5）建立了智能电网环境下家庭能源管理系统实验平台，并对算法进行了初步验证。该平台展示了智能电网环境下，家庭能源管理系统的新特点、新功能和新结构、家庭微电网的构成特点和家庭环境内用户新型用电方式、居民侧用户和电网的互动方式及需求响应实施的效果。利用该平台对本文所提算法进行了初步验证。
英文摘要	The main research work and corresponding contributions of this dissertation can be summarized as follows: (1) This dissertation studies the multi-objective optimization scheduling algorithm for HVAC in smart grid. First, This dissertation presents a new control mode for HVAC in smart grids and introduces a comfort index that is used to quantitatively measure users’ temperature comfort level. Based on this index, a multi-objective scheduling model that simultaneously optimizes electricity cost and temperature comfort level of HVAC users is built. In this model, to cope with the uncertainty due to the error of predicted outdoor temperature, Monte Carlo simulation and scenarios reduction techniques are utilized. The model is resolved by an improved PSO (Particle Swarm Optimization, PSO) algorithm. The effectiveness of the algorithm was verified by simulations. Simulation results show that the algorithm can effectively shift some HVAC load from high price periods to low price periods, which is beneficial to reduce users’ electricity cost and peak load of power system and strengthen the reliability of power system. In addition, the proposed algorithm provides residential users an effective method to take a tradeoff between electricity cost and comfort level conveniently. (2) This dissertation studies algorithm for household microgrid in smart grid, where residential users are electricity prosumers. Firstly, this dissertation proposes a framework of household microgrid and corresponding energy distribution model. The household microgrid framework consists of electricity generation, storage and consumption. In this framework, residential users have the ability to sell surplus electricity to utility grid for revenue. Secondly, according to appliances’ characteristics and user preferences, a set of user comfort level indexes for HVAC, EWH (Electric Water Heater, EWH), EV (Electric Vehicle, EV), WM (Washing Machine, WM), CD (Clothes Dryer, CD), and DW (Dishwasher), which are main controllable appliances in household, are proposed, respectively. Thirdly, based on these indexes, this dissertation builds a multi-objective scheduling model that can simultaneously optimize net electricity cost and comfort level for the proposed household microgrid in smart grids. Finally, the dissertation presents a multi-objective optimization algorithm for HEMS in smart grids and demonstrates its effectiveness by simulations. Simulation results show that the proposed algorithm can effectively flatten residential load by shaving peak and filling valley, take a better tradeoff between electricity cost and comfort level, improve the utilization efficiency of renewable resources, and facilitate the implementation of DR in residential sector. (3) This dissertation studies scheduling algorithm of V2H in the framework of household microgrid in smart grid. This dissertation presents a HEMS framework which has the V2H function and analyses the power distribution relationships among different components in different EV connection statuses and SOC (State of Charge, SOC) scenarios. An optimal scheduling model for HEMS considering the V2H function is built. Finally, this dissertation describes the optimization algorithm with V2H function in detail and demonstrates its effectiveness by simulations. Simulation results show that the V2H function of EV facilitates to reduce residential users’ electricity cost by discharging electricity stored in its battery to supply other loads in household when the utility electricity price is high. (4) Taking multiple EVs’ charging process as an example, this dissertation studies the coordinated optimization algorithm in smart grids to eliminate the load rebound effect caused by a large number of residential users independently scheduling their home appliances under dynamic electricity price mechanism. Firstly, a new coordinated charging mechanism for multiple EVs in smart grids is presented. Following that, the optimal charging scheduling model for each EV is built. Finally, this dissertation describes the coordinated optimal charging algorithm in detail, which is based on water-filling algorithm and proportional charging capacity allocation. The effectiveness of the algorithm is verified by simulations. Simulation results show that the proposed algorithm can effectively avoid overloading distribution transformer. Compared to the current load control mechanism that sending DR signal to specify users’ maximum loads when an overloading event is detected, the coordinated optimal charging algorithm can make more users’ EVs reach the specified SOC values when the scheduling horizon is finished and users pay less for charging. (5) This dissertation builds a HEMS test-bed in smart grids and preliminarily verifies the proposed scheduling algorithms. The test-bed is used to demonstrate the new characteristics, demands and structure of HEMS in mart grids. Through this test-bed, the household microgrid’s characteristics, new electricity consumption mechanism of residential users, and the effects of interaction between residential users and utility grids and DR can be displayed. In addition, the effectiveness of the proposed algorithms are preliminarily verified by the test-bed.
语种	中文
产权排序	1
公开日期	2015-12-27
页码	123页
内容类型	学位论文
源URL	[http://ir.sia.ac.cn/handle/173321/17523]
专题	沈阳自动化研究所_工业控制网络与系统研究室
推荐引用方式 GB/T 7714	张延宇. 智能电网环境下家庭能源管理系统调度算法研究[D]. 中国科学院沈阳自动化研究所. 中国科学院沈阳自动化研究所. 2015.