.png){kind=icon}
.png){kind=icon}
Return
Data-Driven Flexibility Envelopes From AMI for Reliable Grid-Interactive Load Dispatch in Cooling-Dominated Power Systems
.png){kind=icon}
OA
AI
M
M
F
N
DOI：10.1109/access.2026.3720292.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Incorporation of large controllable loads into grid services requires operationally viable flexibility constraints that is operationally viable. Here, empirical flexibility denotes the demand-side operating envelope — hourly capacity bounds, ramp-rate limits, and an energy-neutrality tolerance — inferred directly from measured AMI load variability rather than from assumed or model-based parameters. Existing methods often rely on engineered assumptions or parameters which are obtained using simulation. This paper presents a data-driven approach for deriving empirical flexibility envelopes, namely, ramp-rate limits, hourly capacity limits, and four-hour cumulative-curtailment constraints, derived directly from Advanced Metering Infrastructure (AMI) data. This approach was applied to 300 residential meters in a cooling-dominated environment in Saudi Arabia, with 5.25 million raw 30-minute intervals, and 5.06 million intervals passed a stringent quality-assurance test throughout a one-year period. The integrity of the measurements before the flexibility data extraction is guaranteed by a three-step quality-assurance procedure that includes physics-based validation, temporal-cadence validation, and cross-validation with cumulative energy registers (R<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$2=0.985$ </tex-math></inline-formula>, RMSE = 5.66 kWh/day). The resulting envelopes are strongly seasonally asymmetric: the average load in the hot season is 51% higher than in the mild season, and the flexibility ratios remain stable at 45–55%, indicating that percentage-based constraints are seasonally robust. These empirical parameters are imposed as hard constraints in a closed-loop Model Predictive Control (MPC) simulation of an aggregated-load proxy. Across four 48-h price scenarios, including a case with 20% forecast-error injection, all MPC problems remained feasible and the optimized trajectories satisfied the imposed capacity, ramp, and cumulative-curtailment constraints. Under time-of-use pricing, the controller reduced simulated cost by 22.7% and peak demand by 16.4% relative to the defined simulation baseline. These results establish internal consistency of the constrained simulation; they do not constitute field validation of physical load deliverability. The results indicate that measurement-derived constraints yield constraint-compliant simulated dispatch and reduce the parameter mismatch associated with assumed envelopes. The findings are simulation-based and not a field-deployment validation.
Keywords:
Advanced metering infrastructure
demand-side flexibility
empirical constraint derivation
grid-interactive loads
model predictive control
cooling-dominated power systems
load aggregation
Journal
IF:
3.6
Papers:
9.7W
Citations:
29.4W
Organization
Cited Papers