Original title: Multi-Agent Receding Horizon Games Framework for Autonomous Market Participation
Parth Brahmbhatt, Styliani Avraamidou · August 2026
Small green energy setups like big batteries and hydrogen-producing electrolyzers use huge amounts of power. Buying directly from wholesale electricity markets cuts costs, but these markets demand bid sizes too big for single units to meet. Hiring a middleman solves this, but middlemen charge fees and can favor certain players. Researchers designed a cooperative system where units team up to bid together without a boss. Each device solves a math game to pick its best move, updating every five minutes across a rolling one-hour window. Crucially, units keep their private production and cost data secret, sharing only their total combined power. Testing this on a fleet of six units in the PJM electricity market showed big wins over bidding alone. Every unit ended up better off, gaining 37% higher profit, 29% lower electricity cost, and 92% less wasted, curtailed renewable power.
You and friends plan a road trip together. No one wants to reveal their exact route—it might mean more work. Hiring a tour guide costs money and could favor some. Instead, you only share your planned stops, not the full route, and you adjust every few minutes as traffic changes. Here, energy units share only their total power output, not private data, and re-optimize every five minutes to meet market rules.
For owners of small-scale solar panels, batteries, or hydrogen electrolyzers, joining wholesale electricity markets often requires a minimum bid size they can't meet. Hiring an aggregator adds costs and can be unfair. This research shows that a self-organized peer-to-peer group can achieve 37% higher profits, 29% lower electricity costs, and cut renewable energy waste by 92%, without sharing private data or paying a middleman.
The abstract doesn't say what the limitations are. The method is tested on a six-unit fleet on the PJM market, so results may not apply to other markets or larger groups without further validation.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Multi-Agent Receding Horizon Games Framework for Autonomous Market Participation — Parth Brahmbhatt, Styliani Avraamidou
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.