Electric Vehicle Charging: A Noncooperative Game Using Local Measurements
Electric Vehicle Charging: A Noncooperative Game Using Local Measurements
复制标题
电动汽车充电：使用本地测量的非合作博弈
DOI：
10.3182/20140824-6-za-1003.02072
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
Doreen A. Thomas
中科院分区：
文献类型：
--
作者：
L. Xia;J. Hoog;T. Alpcan;M. Brazil;I. Mareels;Doreen A. Thomas
摘要：提出了一种分布式控制算法，对电动汽车充电时的电力需求进行管理，使总电力需求保持在配电网的限制范围内，并使每辆汽车在充电周期结束时都能获得充分充电的电池。充电算法仅利用局部测量来确定充电电流。将控制问题建模为每个车辆具有弱耦合成本函数的非合作博弈。每辆车的成本函数由单个成本项和组成本项组成。群体成本项表达了所有车辆的总需求，目的是确保基础设施的容量约束得到尊重。结果表明，这一项可以通过局部电压测量来估计。单个成本项反映了在电池中达到期望的充电水平的需要。确定了整个系统存在唯一纳什均衡的充分条件。描述了一种贪婪算法的收敛性和稳定性。为了说明所提出的充电方法的有效性，该算法在澳大利亚郊区低压配电网络的背景下进行了模拟。
Abstract A distributed control algorithm is proposed to manage the electrical power demand for the purpose of charging electric vehicles so that (a) the overall power demand remains within the limitations of the distribution network and (b) each vehicle obtains a sufficiently charged battery at the end of the charging cycle. The charging algorithm utilises only local measurements to determine the charge current. The control problem is modelled as a non-cooperative game with weakly coupled cost functions for each vehicle. The cost function for each vehicle consist of an individual cost term and a group cost term. The group cost term expresses the aggregated demand of all vehicles and serves the purpose of ensuring that the infrastructure capacity constraints are respected. It is shown that this term can be estimated from local voltage measurements. The individual cost term reflects the need to achieve a desired charge level in the battery. Sufficient conditions for the overall system to admit a unique Nash Equilibrium are identified. Convergence and stability properties for a particular greedy algorithm implementation are described. To illustrate the efficacy of the proposed charging methodology, the algorithm is simulated in the context of an Australian suburban low voltage electricity distribution network.