Cloud-edge automated decision support and coordinated control system

Abstract

Active distribution grids require coordinated automated decisions that preserve voltage security while reducing operating CO₂ emissions. However, existing studies generally address carbon-emission-flow tracing, Volt/VAR control, and distributed-energy-resource scheduling separately, leaving a gap between carbon assessment and closed-loop field control. This paper proposes a cloud-edge automated decision-support and coordinated-control system that incorporates nodal carbon intensity into receding-horizon optimization. The cloud performs state estimation, carbon-flow tracing, forecasting, and coordinated scheduling, while edge controllers validate data, enforce release constraints, relay time stamped commands, and provide safe fallback operation. The controller coordinates photovoltaic inverters, batteries, flexible loads, electric-vehicle charging, on-load tap changers, and capacitor banks. A 30-day simulation on a modified IEEE 33-bus feeder demonstrates a 10.6% reduction in operating CO₂ emissions, a 0.62% voltage-deviation index, 2.1% renewable curtailment, a 99.2% command-success rate, and 0.79 s mean end-to-end latency. During a 30 min cloud interruption, edge fallback maintains zero voltage violations and retains a 6.2% emission reduction. These results establish nodal carbon intensity as an actionable and auditable control signal for active distribution-grid operation.