Composite Index for Managing Emergency Repair in Electrical Utilities under Heterogeneous Incident Data

The paper focuses on the management of emergency repair crews in distribution systems characterized by high emergency rates and heterogeneous incident data. The study examines the human-machine system, Human–Tools–Logistics, which necessitates a compact, interpretable, and data-resilient readiness indicator. The scientific novelty lies in the development of a non-compensatory composite operational readiness index for crews, based on the geometric aggregation of three sub-indices: Human, Tools, and Logistics. A two-stage methodology for constructing the composite operational readiness index is proposed. In the first stage, heterogeneous data and expert knowledge regarding personnel, equipment, logistics, and environmental conditions are integrated through a multi-agent data fusion framework. In the second stage, normalized sub-indices (H/T/L) are developed using fuzzy logic. Finally, the operational readiness index is aggregated as a normalized weighted geometric mean, ensuring its non-compensatory nature. Simulation software was developed to model emergency scenarios in Moscow power grids. For a fixed point in time, the index values are calculated for multiple crews to match them with an incident and establish the most effective crew for that incident. The impact of temperature, traffic congestion, crew size, along with personnel experience and qualifications is analyzed. The analysis demonstrates diminishing returns on crew size and saturation of the experience curve, alongside the significant influence of adverse meteorological conditions and traffic congestion. The proposed index is compared with existing simulation-based, probabilistic, and optimization-oriented models, as well as fuzzy-logic and decision support systems. The findings demonstrate that the proposed operational readiness index offers superior interpretability for dispatcher decision-making and can be seamlessly integrated into assignment and routing optimization tasks, facilitating bottleneck identification and dispatcher training.