【作者】 李秋硕；

【导师】 肖湘宁；

【作者基本信息】 华北电力大学 ， 电力系统及其自动化， 2014， 博士

【摘要】 电动汽车是世界能源技术革命和国家新能源战略的重要组成部分,是国家七大战略新兴产业之一。未来规模化的电动汽车充电将给电网的运行带来深远的影响和挑战,新能源接入、电力系统安全经济运行与电动汽车充放电三者之间的相互作用和关系,是新能源电网和电动汽车发展面临的重要问题。本文围绕电动汽车接入电网后的电能有序利用问题展开研究和分析,主要成果和创新点如下：根据对可控热负荷控制思想和控制策略的分析,提出了可控负荷需具备的两个必要条件：1)功率需求定义的时间尺度远大于电力系统工频周期,2)负荷允许的最大工作功率大于平均功率需求；基于上述讨论和不同类型电动汽车行泊规律和充电行为的统计分析,论证了电动汽车负荷具备可控性,揭示了充电负荷具备灵活性和可调节性的物理本质和基本原理,明确了有序充放电的控制对象。充电需求是充放电功率有序控制的基础和重要约束。基于动力电池的戴维南等效电路模型,建立其充放电过程的稳态功率模型,描述不同电池状态、不同充电方式下的充电功率和电池电能的累积过程；在充电功率可变的情况下,考虑到电池的最大受电能力和充电机的最大充电功率的约束,建立了单一电动汽车的充电需求模型,反映用户的充电时间和能量需求以及充电功率的约束；针对动力电池充电过程的功率具有有后效性的特点,创造性地提出应用时间序列分析理论预测未来某区域内电动汽车充电站桩的投入数量,建立了多辆电动汽车集中充电负荷的充电需求预测模型,为充放电功率的有序控制确定了描述能量和时间约束的可行域边界。参与电网有功和能量调度是电动汽车有序充放电最主要目标。在有功能量调度方面,提出含新能源接入的有源城市电网规模化电动汽车的充电功率有序控制策略。构建了双层控制结构,阐述了两个控制层面间的相互约束关系,实现了充电功率规划与瞬时功率的解耦控制；在电网层,以提高新能源利用率、平抑配电网的有功负荷波动和火力发电机组经济调度为目标,针对电动汽车的特点,考虑用户的充电能量、时间需求约束和电池、充电机的充电功率约束,提出了充电功率多目标优化模型,和求解该约束多目标问题的约束多目标差分进化算法,并应用基于信息熵权的逼近理想解排序多属性决策方法从Pareto前沿中选取最优解,获得日前最优充电功率规划；在用户层,基于电池的受电能力特性和用户的心理承受力,提出分段加权功率分配方法,对不同充电阶段的电池阶段赋予不同的充电优先级和不同的功率分配策略,满足用户的充电能量需求和功率分配公平性的要求,同时最大化电网层优化模型中最大充电功率约束的范围。利用充电设施的无功补偿能力,辅助电网无功/电压控制。提出利用非车载快速充电机作为无功电源,实现无功就地补偿的模型和优化算法。考虑充电负荷的特点带来的无功调节范围约束,建立以网损最小化为目标的无功优化模型；应用拉格朗日函数和梯度法,提出初始可行解的搜索方法,获得满足节点电压偏差要求的初解；将非线性问题线性化后,讨论步长对求解精度的影响,并给出步长的确定方法；求解线性规划问题,寻找满足网损最小化目标的最优解,实现配电网无功/电压控制。面向靠近变电站、为电动私家车和出租车提供充电服务的集中型电动汽车充换电站,在现有充电站通讯和监控系统基础上,基于Web平台研究开发了有序充电控制软件。在SQL Server数据库系统下建立全站的数据采集、管理与分析的基础数据平台；对充电功率控制策略进行适当改进,并在C#环境下编写程序智能调度电动汽车的充电行为,实现有序充电；软件还集成了充电站信息管理、智能引导与可视化监控等功能,实现充电站管理与电能使用的有序化、智能化、可视化。更多还原

【Abstract】 Electric Vehicle (EV) is an important part of the world’s energy technology revolution and the national renewable energy strategy. It is one of the seven Emerging strategic acquisitions in China. Scales of EV charging will effect the power system significantly. Interaction of integration of renewable energy, security and economic of power system operation, charging and discharging of EVs is an important problem in development of EV and active power system. Concentrating on coordinated consumption of electric power energy with EV intergrated, this paper has conducted the following research:According to the control strategy of thermal load, the idea of average power value in large time scale is not a strict constraint to the average power in small time scale is proposed first. Applying this idea, based on statistics and analysis of driving and charging behaviors of different kinds of EVs, physical reality and mechanism of flexibility and adjustability of EV charging load are revealed. Controlling objective of coordinated charging and discharging is pinpointed.A steady-state model is established based on Thevenin equivalent circuit of EV battery. Charging power under different state of charge (SOC) and charging mode and energy accumulate process is discribed. A single EV charging demand model is established considering maximum charging power of battery and charger under variable charging power so that users’demand on energy and charging time and constraints of charging power can be reflected precisely. As EV charging is a follow-up-effect process, charging behavior forecast approach is proposed based on time sequence analysis theory. And a model of charging demand forecast of scales of EVs in an area is established. This work provides feasible region constraint to charging power planning.A coordinated EV charging power control strategy in active city power system is proposed. A double-layer control system is presented. Constraints between two control layers are revealed. And decoupling of charging power planning and real-time power control is achieved. On the system layer, aiming at maximum utilization rate of renewable energy, stabilizing load fluctuation and economic dispatch of thermal generators, considering power system operation and charging demand constraints, a muti-objective optimazation model is established. A nevel constrained multi-objective differential evolution (CMODE) algorithm is proposed to get the pareto solution of the multi-objective optimazation problem. A comentropy weighted technique for order preference by similarity to an ideal solution decision-making method is applied to select optimal solution from pareto solutions, so that the optimal day-ahead charging plan is obtained. On the user layer, according to analysis of battery charging power and users’ psychological enduring capacity, a piecewise weighted power dispatching (PWPD) approach is proposed. Different priority and power dispatch strategies are employed in the light of battery state and charging period. As a result, charging demand can be met, fairness among users are satisfied, and EVs can provide a maximum adjustable power to the system layer control.In microgrids, a sheme of local reactive power compensation utilizing off-board fast chargers is put forward. Considering reactive power compensation capacity of the chargers, the objective of var/voltage control (VVC) is formulated as minimizing system power loss while regulating voltage profile within acceptable limits. The centralized VVC scheme is a two-phase control scheme where gradient based Phase I control targets at regulating voltages and linear programming (LP)-based Phase II optimization aims at minimizing power loss. Effects of step length caused by linearization is discussed, and the way to determine step length in linear programming is given.A coordinated charging control system based on Web is designed and developed in this paper. This system serves for charging station which is near the substation and charge for priviate electric cars and electric taxies. Data platform including collection, administration and analysis is built in SQL Server. Control strategy of coordinated charging is modified to adapt the situation of charging station. Smart control algorithm program are coded in C#. EVs charging in the station can be controlled by the program, so that coordinated charging is realized. Besides, this system also integrates imformation management, smart guiding and visual monitoring. Coordinated, smart, visual consumption of electric power in EV charging stations is achieved.更多还原