Dynamic Flexibility Duration Assessment of Low-Voltage Feeders Using Multi-Feature Trajectory Similarity Matching

Abstract

Increasing distributed energy resource penetration makes low-voltage feeder flexibility highly time-varying and uncertain, limiting the operational value of static capacity-based estimates. This paper proposes a dynamic flexibility duration assessment framework using multi-feature trajectory similarity matching. Historical operating periods are matched using temporal, calendar, load-state, weather-state, and photovoltaic-state features to construct plausible future feeder trajectories. Metadata-constrained models then estimate upward and downward flexibility from photovoltaic systems, electric vehicle chargers, heat pumps, and battery storage, while feeder-level operating constraints and duration-dependent battery energy-feasibility checks are applied to maintain physical consistency over the requested service interval. The resulting flexibility trajectories are transformed into a future flexibility duration matrix that links flexibility magnitude, continuous service duration, and dependable level through lower-tail ensemble quantiles. Case studies based on realistic minute-level feeder data examine peak-load, photovoltaic-rich, and transition operating conditions. Under the examined ablation setting, the proposed matching method achieves lower trajectory errors than the compared similarity strategies, and the resulting assessments reveal distinct upward-dominant, downward-dominant, and bidirectional flexibility characteristics across the investigated scenarios. Parameter sensitivity analysis indicates local robustness under the tested MTSM configurations, while a cross-feeder evaluation using 20 representative feeders examines applicability beyond a single feeder. Compared with static assessment, the proposed dynamic approach provides more conservative but operationally meaningful flexibility estimates by accounting for future operating uncertainty, time-varying resource availability, and sustained service requirements. The framework offers a practical means of quantifying duration- and dependability-aware feeder-level flexibility under partial observability.