Research & Papers

FP-AMM algorithm cuts grid fairness error by 55% on IEEE benchmarks

A stateful market maker that guarantees fair electricity allocation even during scarcity events.

Deep Dive

The Fair Play Automatic Market Maker (FP-AMM) introduces a stateful, auditable approach to electricity scarcity allocation, addressing the memoryless nature of traditional locational marginal pricing. The mechanism uses a two-stage stochastic clearing rule—service-priority sampling plus inverse-fairness weighting—combined with a DC-OPF feasibility set and a bounded shortage memory that updates via a saturated integrator. This design guarantees per-node delivery ratios converge almost surely to contracted fairness targets, with theoretical finite-time O(1/√T) bounds on deviation.

Validated on IEEE 14-, 57-, and 118-bus networks over 5,000 market intervals, FP-AMM achieved convergence to fairness targets across all benchmarks. During scarcity periods, the mechanism cut peak weak-bus fairness error by 54% on the IEEE-57 network and up to 55% compared to an equal-weight baseline. The system maintained DC feasibility throughout, and event-triggered execution ensured practical boundedness of allocation tracking error with quantified computation-fidelity trade-offs. This work establishes rigorous convergence proofs, including invariant shortage-memory states and linear contraction rates for the clearing operator.

Key Points
  • FP-AMM uses a two-stage stochastic clearing with service-priority sampling and inverse-fairness weighting tied to a DC-OPF feasibility set.
  • Fairness error reduced by 54% on IEEE-57 network and up to 55% relative to equal-weight baseline during scarcity.
  • Mechanism converges almost surely to contracted fairness target with finite-time O(1/√T) bound via Lyapunov analysis.

Why It Matters

A practical algorithm to ensure equitable power distribution during grid stress, crucial for decarbonized grids with variable renewables.

📬 Get the top 10 AI stories daily