A multi-microgrid system, including several microgrids and distributed energy resources, is always threatened by numbers of faults and attacks as a consequence of which malfunctioning can occur on a large scale. Thus, minimizing the effects of such disruptions is of paramount importance. This paper addresses the problem of mitigating a multi-microgrid system that faces false data injection and replay attacks by considering the multi-microgrid as a multi-agent system in which each microgrid as an agent represents a node in a weighted directed graph. The problem of consensus among normal agents is studied when microgrids and their communications are attacked. The malicious agents become isolated with the help of Weighted Mean Subsequence Reduced (W-MSR) algorithms in which all normal agents neglect the extreme values received from their neighbors. The proposed controller is able to maintain the system’s desired performance when false data is injected into the system, or valid data is received with time-delays. Finally, numerical examples and simulation results are provided.

包含多个微电网和分布式能源的多微电网系统始终受到大量故障和攻击的威胁，其结果是可能会大量发生故障。因此，使这种干扰的影响最小化是至关重要的。本文通过将多微电网视为多智能体系统来解决缓解面临错误数据注入和重播攻击的多微电网系统的问题，其中，每个微网作为智能体，代表加权有向图中的一个节点。研究了当微电网及其通信受到攻击时，正常代理之间的共识问题。借助降低的加权平均子序列（W-MSR），隔离了恶意代理所有正常代理都忽略了从邻居那里收到的极值的算法。当将错误的数据注入到系统中或接收到具有时延的有效数据时，建议的控制器能够保持系统所需的性能。最后，提供了算例和仿真结果。