Fractal and Markov-Based Load Modeling for Multi-Generator Dispatch and Protection in Urban Networks

Abstract This paper presents a unified simulation framework for urban multi-generator networks that integrates stochastic load modeling, predictive health assessment, and protective mechanisms. Consumer demand is represented by a Markov chain augmented with fractal noise to capture both state-based transitions and long-range correlations in usage. Generators are coupled with digital twin models employing non-homogeneous Poisson processes, enriched with harmonic resonance factors, to forecast failure risk and dynamically derate available capacity. A closed-loop controller anticipates total system demand, allocates generation proportionally to health-adjusted capacity, and enforces fuse and harmonic-trip logic to mitigate overloads and cascading failures. Simulation results over 24-hour horizons demonstrate realistic peak–valley patterns, effective cluster-based demand segmentation, and robust balancing of generation with zero unexpected protective trips.