面向新能源智能电网的多Agent信息过滤网络安全增强方法

A MULTI-AGENT INFORMATION FILTERING SECURITY ENHANCEMENT METHOD FOR NEW ENERGY SMART GRID

针对新能源智能电网的电力系统状态估计问题,提出一种基于多Agent信息过滤的智能电网安全增强方法,以保护系统状态估计过程不受来自故障设备或网络攻击的错误数据注入的影响。基于多Agent过滤方案进行电网状态估计,各Agent采用基于信任机制的过滤方案,将信任度系数作为所接收到的对应邻接Agent状态值的权重参数。同时,考虑到每个Agent的信任度动态变化的可能性,设计实现Agent信任度更新机制,将Agent的长期行为特征作为量化其信任度的依据。在模拟电网中的仿真实验验证了该方法的有效性,能够应对错误数据注入对电网状态估计的影响。

Aiming at the problem of power system state estimation in new energy smart grid,this paper proposes a security enhancement method for smart grid based on multi-agent information filtering to protect the system state estimation process from the influence of wrong data injection from faulty equipment or network attacks.The power grid state estimation was performed based on a multi-agent filtering scheme.Each agent adopted the filtering scheme with a trust-based mechanism.It took the trust coefficient as the weight parameter of the received corresponding adjacent agent state value.Considering the possibility of the dynamic change of the trust degree of each agent,the agent trust degree update mechanism was designed and implemented,and the long-term behavior characteristics of the agent were taken as the basis to quantify its trust degree.The effectiveness of this method was verified by a simulation experiment of power grid,which could effectively deal with the influence of false data injection on power grid state estimation.