The accelerating global energy transition, driven by climate imperatives and technological advancements, demands fundamental transformations in power systems. Smart grids, characterized by cyber–physical integration, distributed renewable resources, and data-driven intelligence, have emerged as the backbone of this evolution. This convergence, however, introduces unprecedented complexities in resilience, security, stability, and market operation. This special issue presents five pivotal studies addressing these interconnected challenges, offering novel methodologies and insights to advance the efficiency, resilience, and sustainability of modern power systems.

Among these critical challenges, the increasing frequency and intensity of hazardous weather events pose a significant threat to the power infrastructure. In this issue, Zhiyi Li et al. present a nuanced classification of weather threats, ranging from short-lived extreme events such as hurricanes and ice storms to prolonged adverse conditions such as renewable energy droughts during monsoon seasons. Their analysis reveals how these threats trigger cascading impacts across both physical and cyber domains. For instance, extreme winds can damage overhead lines and communication cables simultaneously, while long-term temperature anomalies degrade the dynamic thermal ratings of power lines and disrupt wireless communication through atmospheric duct effects. These interdependencies lead to common-cause failures (e.g., substation outages disabling connected remote terminal units (RTUs)) and cascading failures (e.g., communication breakdowns paralyzing situational awareness). To address these challenges, the authors advocate for a comprehensive resilience cycle that encompasses cyber–physical collaboration. This includes leveraging advanced sensing for real-time hazard assessment, proactive resource prepositioning informed by digital twins, and decentralized control during communication degradation. Their framework culminates in a quantitative resilience evaluation that integrates cyber-specific metrics such as observability loss and controllability degradation, offering a holistic blueprint for increasing active distribution network (ADN) resilience in an era of climate uncertainty.

Beyond addressing physical and cyber resilience to environmental threats, the low-carbon transition itself necessitates innovative market mechanisms. The rise of “prosumers” introduces bidirectional power flows and complex challenges in carbon responsibility allocation. Ma et al. develop a sophisticated carbon-coupled network charge-guided bi-level interactive optimization framework. The upper level, managed by the distribution system operator (DSO), formulates carbon-coupled network charges by integrating a carbon-emission flow model with optimal power flow, explicitly accounting for peer-to-peer (P2P) trading impacts. The lower level facilitates decentralized P2P market clearing for both energy and carbon-emission rights among prosumers. An adaptive alternating direction method of multipliers (ADMM) combined with an improved bisection method ensures efficient convergence to market equilibrium. Tested on a modified Institute of Electrical and Electronics Engineers (IEEE) 33-bus system, this model provides a blueprint for harmonizing economic efficiency, network security, and carbon-reduction goals in future distribution networks. The framework not only addresses technical aspects of power system operation but also integrates economic and environmental considerations, offering a comprehensive approach to managing modern power system complexities.

The stability of these evolving power systems is further complicated by the widespread integration of power electronic devices, which render traditional phasor-based analysis inadequate when node voltages exhibit time-varying amplitude and frequency (TVAF) characteristics. Hu et al. provide a fundamental redefinition of power response analysis for power-electronics-dominated grids. Through a rigorous mathematical derivation based on original network relationships and superimposed step responses, they reveal complex multi-timescale dynamics in active and reactive power flows. These dynamics arise from the interplay of voltage amplitude oscillations, frequency fluctuations, and their harmonic interactions. The authors uncover novel phenomena such as network power storage and release, directly challenging the conventional view that inductors merely transfer active power. By deriving current responses using a superimposed step-response methodology and rigorous Taylor series expansions, they demonstrate how rapid voltage dynamics generate broadband harmonic components across multiple orders. For practical application, Hu et al. establish convergence boundaries for series simplification and provide engineering-viable expressions that retain dominant dynamics while accommodating real-world constraints. Empirically validated using field data from renewable generation bases, this work offers indispensable insights for stability assessment and control in grids dominated by inverter-based resources.

However, the sophisticated cyber infrastructure enabling ADN resilience and control also introduces significant new vulnerabilities. As distributed power supplies proliferate, so too do the attack surfaces for malicious actors. Liu et al. introduce a novel and potent black-box false data injection attack (FDIA) method. Unlike traditional attacks targeting communication networks, this approach directly compromises the measurement modules within distributed power supplies. By employing a generative adversarial network (GAN) to generate stealthy attack vectors, it bypasses conventional security measures without requiring detailed system knowledge. This research starkly highlights the evolving cybersecurity risks inherent in the data-driven algorithms that increasingly govern smart grids, demanding urgent innovation in defensive strategies to protect system stability. The authors emphasize that attackers can now exploit previously overlooked entry points, such as measurement modules, to inject false data and disrupt operations. This necessitates a reevaluation of traditional security measures and the development of more robust defenses to counteract such sophisticated cyber threats.

Ensuring transient stability within these complex grids, particularly under disruption threats, requires rapid and accurate assessment tools. Leveraging the differential topological properties of the dynamic security region (DSR)—notably its hole-free interior and tight boundaries—Liu and Jia propose an innovative two-stage method. First, a space division technique drastically compresses the search space to locate the critical operation area housing the practical DSR (PDSR) boundary. Second, a Wasserstein generative adversarial network with a gradient penalty (WGAN-GP) rapidly generates a large number of critical points based on a small training set derived from the compressed space. The hyperplane-based PDSR boundary is then efficiently fitted using least squares. This method is a prime example of the integration of classic power system models with advanced artificial intelligence (AI) techniques. The space division approach, rooted in traditional power system analysis, works in tandem with the WGAN-GP model, a state-of-the-art machine learning method. Their combination not only preserves the physical interpretability of classic models but also increases the efficiency and adaptability of the assessment process through AI’s superior pattern-recognition capabilities. Validated on the IEEE 39-bus system, this hybrid approach provides a powerful tool for real-time situational awareness and security-constrained optimization, enabling more efficient and accurate power system stability assessments.

The studies in this special issue collectively advance smart grid resilience, security, stability, and sustainability. They address critical challenges—from hardening cyber–physical systems against natural disasters and stealthy false data injection attacks, to deciphering power dynamics under time-varying excitations and enabling rapid stability assessment via AI-driven boundary generation, and, finally, to pioneering market mechanisms that align peer-to-peer energy trading with carbon accountability. These advances emphasize that solving modern power system complexities demands not only domain-specific innovations but also deep integration across cyber–physical layers and techno-economic dimensions.

The novel methodologies presented in this issue—spanning GAN-based attack simulations, TVAF response analysis, space-division-accelerated security regions, and carbon-coupled market optimization—provide indispensable tools for future grid design. As the energy transition intensifies, sustained interdisciplinary collaboration across engineering, computer science, and economics will be paramount. We sincerely thank the authors, reviewers, and editorial team for their vital roles in assembling this timely contribution toward building resilient, secure, and sustainable smart grids.