In scenarios such as disaster response, remote area surveillance, and military reconnaissance, where stable communication infrastructure may be unavailable, unmanned aerial vehicles (UAVs) are often deployed to physically carry data. Ensuring data reliability in such scenarios, especially in adversarial environments, requires data backup mechanisms for UAV swarms. This paper addresses the critical yet often overlooked challenge of data recovery difficulty in UAV swarm networks. We introduce Data Recovery Entropy (DRE) to quantify the spatial dispersal of backed-up data across the swarm. Based on this metric, we propose two data backup strategies: one enables adjustable level of DRE through a tunable parameter, and the other prioritizes minimizing DRE. The proposed data backup strategies are tailored to the unique characteristics of UAV swarms, focusing particularly on their dynamic communication topologies. Finally, we vary system parameters, including node deployment parameters, communication parameters, and mobility parameters, to validate the proposed approach.

In this paper, we demonstrate the interdependence among buses in power systems that experience voltage instability events by deriving interaction graphs through successive transition probability matrices and the Markov transition matrix. A quasi-stationary convergence has been observed in the probability transition matrices, suggesting a pattern in voltage instability failure. We propose algorithms to identify the pattern, referred to as fault chains, by tracing the interaction graphs. The interaction graphs and fault chains can be used to formulate strategies for enhancing network robustness, including the assignment of new photovoltaic buses, load rescheduling, and equipping droop controllers. Numerical experiments performed on the IEEE 118- and 300-bus systems demonstrate that equipping droop controllers at selected buses is an effective approach for improving network robustness.

In this paper, we investigate the influence of the locations of inverter-based resources (IBRs) on the synchronization performance of power networks. We propose two indexes to measure the distribution of inertia in power networks, considering the distribution of control parameters and the topological factors jointly. The first index is the inertia clustering coefficient, which measures how densely the neighbors of each node are connected in a power network. The second index is the inertia centrality coefficient, which captures whether the distribution of inertia in a power network is centralized or peripheral. We characterize synchronous generators (SGs), grid-following inverters (GFLs), and grid-forming inverters (GFMs) by their damping and inertial properties. We evaluate the synchronization performance of the system after disturbances by adopting the settling time and hertz-sec metric. Both the frequency response and the trajectories of eigenvalues show that the location of IBRs has a significant impact on the synchronization performance of the system. Monte Carlo simulations are conducted on two test networks, the results of the IEEE 30-, 57-, and 118-bus systems demonstrate a strong correlation between synchronization performance and the two inertia distribution indexes.

This paper studies cascading failure propagation in power systems and presents methods for the quantification of important properties of failure propagation. First, the topological properties of cascading failure propagation are examined. This includes an analysis of the electrical distance between consecutive failures, shedding light on the spatial spread of failures. Additionally, the formation of islands in cascading failure processes is explored to understand their topological characteristics. Second, the paper measures the evolution of a power system during cascading failure processes, considering both the topological changes of the overall system and the propagation rates of system loss. This analysis provides a measurable comprehension of how the system evolves and adapts as failures propagate. Third, this study investigates system loss and analyzes the contributions of various failure mechanisms to the overall system loss. Numerical experiments yield valuable insights into the propagation of cascading failures, leading to several significant conclusions. The findings from this research can inform the development of effective strategies for resilience enhancement and risk mitigation.

Power systems around the world exhibit diverse topological characteristics depending on historical and geographical factors. In this paper, we aim to acquire insights into the relationship between the topological characteristics of power systems and their robustness through controlled experiments. To achieve this, we generate synthetic networks that incorporate the generation and load settings of practical power systems. We test various aspects, such as topological properties of regional power grids, interconnections among regional power grids, network sizes, and power balance of regional power grids. We derive inspiring results by performing cascading failure trials in test networks. Specifically, a wide and compact power grid can be more robust than a narrow and elongated power grid. Moreover, increasing the capacity of interconnection links among regional power grids may result in a less robust power system concerning cascading failures.

This study proposes a cascading failure model for power systems that addresses event-triggered power flow divergence. After a cascading failure event, voltage instability is a major cause of event-triggered power flow divergence. To detect voltage instability events, we track the steady-state voltage profiles across successive cascade generations by adopting continuation power flow (CPF). Various cascading failure events are incorporated into the CPF computation. Hence, the proposed model can identify voltage instability by monitoring the emergence of buses that hit a saddle-node bifurcation (SNB) point. In this way, the proposed cascading failure model is able to derive equilibrium points, if they exist, under all conditions. Furthermore, the model incorporates primary frequency control instead of slack buses to account for power losses, thereby improving the accuracy of the results. In addition, we propose a set of metrics to measure the cascading failure propagation rate and the influence of voltage instability events on cascading failure outcomes. Experimental results indicate that power system configurations exert a significant influence on voltage instability events.

Some components in a power system are more vulnerable than others. Limited by the capability of the existing quasi-steady-state (QSS) models of cascading failures, previous analysis on vulnerable components typically omits the influence of voltage collapse. In this paper, we identify the nodes that are most susceptible to voltage collapse in cascading failures with a QSS model that applies practical measures in response to voltage collapse, i.e., undervoltage load shedding (UVLS) and protection relays. The proposed model is based on the continuation power flow (CPF). We apply the model to the IEEE 118-bus system, the IEEE 300-bus system, and the Polish 2383 bus system. We count the times and plot the locations of buses that exhibit a voltage collapse event in the network to assess structural vulnerability. Numerical results show that some parts of the network are more vulnerable to voltage collapse events than others.