The transition towards increased utilization of renewable energy and electric vehicles (EVs), along with the growing use of various other electrical devices, poses challenges to the stable and resilient operation of electric power systems (EPS), especially in the face of natural phenomena associated with climate change. This means that accurate topology and balanced EPS plays a key role to increase the capacity to respond quickly and in a coordinated manner to disaster situations such as cyber-attacks, earthquakes and floods. In this study, a new approach is presented to quickly and accurately detect topology attacks in EPS, thus contributing to making safer and more resilient. The proposed methods provide insights into maintaining uninterrupted electricity service by enabling EPS management through both post- and pre-event operational strategies. This approach is created by identifying faulty points with the obtained topology information and creating microgrid (MG) groups. Machine learning techniques have been integrated into the data intrusion attack detection (DIAD) system, enabling the detection of manipulated or faulty smart meters (SM). Concurrently, a topology identification (TI)-based graph learning algorithm is propounded to determine the exact fault locations before and after the event. For MV region restoration after determining the TI region, a mixed-integer linear programming (MILP) approach is employed to optimize the load restoration process in the MG regions. This approach aims to minimize losses and restore critical loads to their previous state as quickly as possible using flexible and emergency power balancing systems, including grid-support storage systems (GSSs), photovoltaic systems (PVs), electric vehicle charging stations (EVCS), and mobile generators. Moreover, a detailed compilation is presented under the topics of EPS topology, phase identification (PI) and its effect on power system resiliency (PSR), shedding light on the future development of EPS.
