Reactive power and voltage control across electrical systems: A comprehensive review of methods, challenges, and future directions

The increasing penetration of inverter-based resources (IBRs) and distributed energy resources (DERs) is making reactive-power and voltage control (RPVC) increasingly complex across transmission, distribution, and microgrid systems. Existing RPVC approaches span diverse optimization, control, and learning paradigms, making their operational capabilities and remaining deployment gaps difficult to compare consistently. This article presents a PRISMA-guided systematic review complemented by systematic cross-axis evidence mapping. Database searching and supplementary citation tracking identified 678 records; after deduplication, screening, eligibility, and technical-quality assessment, 168 unique references were retained. A three-axis taxonomy classifies RPVC studies by control objective, implementation architecture, and temporal scope, while application domain is retained as a contextual attribute. Of the retained literature, 65 eligible primary studies provided sufficient information for consistent quantitative coding across all three facets. The mapping shows that online/real-time operation is the largest temporal category (31 studies, 47.7%), centralized implementation the largest architectural category (27, 41.5%), and multi-objective operation the largest objective category (26, 40.0%). Decentralized/local online/real-time control (13 studies) and hierarchical/hybrid multi-timescale coordination (12 studies) are well represented, indicating that real-time or distributed operation alone is not a general research gap. In contrast, only one mapped study combines an explicit stability/security objective, distributed/multi-agent architecture, and online/real-time operation. A common evidence matrix further indicates that the reviewed approaches do not generally demonstrate certified global optimality, explicit discrete-device fidelity, and sub-minute operational capability simultaneously. The combined evidence shows that the more consequential research needs lie at technically demanding intersections involving stability and feasibility assurance, cyber-resilient distributed coordination, constraint-safe learning-enabled control, and coordination of heterogeneous devices across timescales. The review therefore provides an evidence-based framework for comparing RPVC implementations and prioritizing stability-aware, constraint-safe, cyber-resilient, privacy-preserving, and edge-enabled control.