A Risk-Based Optimization Model for Electric Vehicle Infrastructure Response to Cyber Attacks
A Risk-Based Optimization Model for Electric Vehicle Infrastructure Response to Cyber Attacks
复制标题
DOI：
10.1109/tsg.2017.2705188
复制
发表时间：
2018-11
影响因子：
9.6
通讯作者：
S. Mousavian;M. Erol-Kantarci;Lei Wu;T. Ortmeyer
中科院分区：
文献类型：
--
作者：
S. Mousavian;M. Erol-Kantarci;Lei Wu;T. Ortmeyer
当电动汽车（EV）基础设施的脆弱性没有得到适当解决时，智能电网的安全性就面临风险。随着汽车变得越来越智能和互联，它们被入侵的风险也在增加。在各种场合和场所，黑客入侵智能或自动驾驶汽车已被证明是可能的。在大多数情况下，受损车辆对驾驶员和其他车辆构成威胁。另一方面，当车辆是电动的时，攻击可能从EV供电设备（EVSE）开始一直蔓延到公用事业系统的电网基础设施。传统的基于隔离的保护方案在智能电网中不能很好地工作，因为电力服务具有可用性限制，并且很少有组件具有物理备份。在本文中，我们提出了一个混合整数线性规划模型，共同优化安全风险和设备的可用性，在相互依赖的电力和电动汽车基础设施。我们采用流行病攻击模型来模拟恶意软件的传播。我们假设恶意软件在EV充电期间传播，当EV从受感染的EVSE充电，然后在另一个EVSE上旅行和充电时。此外，它还通过EVSE的通信网络传播。所提出的响应模型旨在隔离受损和可能受损的EVSE的子集。响应模型最大限度地减少了攻击传播的风险，同时提供了一个令人满意的设备供应需求的可用性水平。我们的分析表明，在智能电网中的电动汽车基础设施的攻击所提出的响应模型的理论和实践的界限。
Security of the smart grid is at risk when the vulnerabilities of the electric vehicle (EV) infrastructure is not addressed properly. As vehicles are becoming smarter and connected, their risk of being compromised is increasing. On various occasions and venues, hacking into smart or autonomous vehicles have been shown to be possible. For most of the time, a compromised vehicle poses a threat to the driver and other vehicles. On the other hand, when the vehicle is electric, the attack may spread to the power grid infrastructure starting from the EV supply equipment (EVSE) all the way up to the utility systems. Traditional isolation-based protection schemes do not work well in smart grid since electricity services have availability constraints and few of the components have physical backups. In this paper, we propose a mixed integer linear programming model that jointly optimizes security risk and equipment availability in the interdependent power and EV infrastructure. We adopt an epidemic attack model to mimic malware propagation. We assume malware spreads during EV charging when an EV is charged from an infected EVSE and then travels and recharges at another EVSE. In addition, it spreads through the communication network of EVSEs. The proposed response model aims to isolate a subset of compromised and likely compromised EVSEs. The response model minimizes the risk of attack propagation while providing a satisfactory level of equipment availability to supply demand. Our analysis shows the theoretical and practical bounds for the proposed response model in smart grid in the face of attacks to the EV infrastructure.