Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Reinforcement Learning Enabled Peer-to-Peer Energy Trading for Dairy Farms

Mian Ibad Ali Shah, Enda Barrett, Karl Mason

TL;DR

The paper addresses high energy costs and CO2 emissions in dairy farms by leveraging renewable generation and peer-to-peer energy trading within a multi-agent system. It presents MAPDES, a simulator that integrates rule-based agents with a Q-learning agent in a distributed P2P trading market governed by a Double Auction and SDR pricing. The study demonstrates substantial improvements over baselines, including a 70.44% reduction in electricity purchases, an 87.84% drop in peak grid demand, and a 1.91% increase in revenue from surplus energy, validating the value of combining reinforcement learning with rule-based control for agricultural energy management. The work highlights the potential for scalable, privacy-preserving energy optimization in dairy farming and suggests extending the approach to other agricultural domains."

Abstract

Farm businesses are increasingly adopting renewables to enhance energy efficiency and reduce reliance on fossil fuels and the grid. This shift aims to decrease dairy farms' dependence on traditional electricity grids by enabling the sale of surplus renewable energy in Peer-to-Peer markets. However, the dynamic nature of farm communities poses challenges, requiring specialized algorithms for P2P energy trading. To address this, the Multi-Agent Peer-to-Peer Dairy Farm Energy Simulator (MAPDES) has been developed, providing a platform to experiment with Reinforcement Learning techniques. The simulations demonstrate significant cost savings, including a 43% reduction in electricity expenses, a 42% decrease in peak demand, and a 1.91% increase in energy sales compared to baseline scenarios lacking peer-to-peer energy trading or renewable energy sources.

Reinforcement Learning Enabled Peer-to-Peer Energy Trading for Dairy Farms

TL;DR

The paper addresses high energy costs and CO2 emissions in dairy farms by leveraging renewable generation and peer-to-peer energy trading within a multi-agent system. It presents MAPDES, a simulator that integrates rule-based agents with a Q-learning agent in a distributed P2P trading market governed by a Double Auction and SDR pricing. The study demonstrates substantial improvements over baselines, including a 70.44% reduction in electricity purchases, an 87.84% drop in peak grid demand, and a 1.91% increase in revenue from surplus energy, validating the value of combining reinforcement learning with rule-based control for agricultural energy management. The work highlights the potential for scalable, privacy-preserving energy optimization in dairy farming and suggests extending the approach to other agricultural domains."

Abstract

Farm businesses are increasingly adopting renewables to enhance energy efficiency and reduce reliance on fossil fuels and the grid. This shift aims to decrease dairy farms' dependence on traditional electricity grids by enabling the sale of surplus renewable energy in Peer-to-Peer markets. However, the dynamic nature of farm communities poses challenges, requiring specialized algorithms for P2P energy trading. To address this, the Multi-Agent Peer-to-Peer Dairy Farm Energy Simulator (MAPDES) has been developed, providing a platform to experiment with Reinforcement Learning techniques. The simulations demonstrate significant cost savings, including a 43% reduction in electricity expenses, a 42% decrease in peak demand, and a 1.91% increase in energy sales compared to baseline scenarios lacking peer-to-peer energy trading or renewable energy sources.
Paper Structure (17 sections, 1 equation, 6 figures, 1 table)

This paper contains 17 sections, 1 equation, 6 figures, 1 table.

Table of Contents

  1. Introduction
  2. Centralized Trading
  3. Decentralized Trading
  4. Distributed Trading
  5. Our Approach
  6. Background
  7. Rule-based Systems
  8. Reinforcement Learning
  9. Q-learning
  10. System Design
  11. Dataset and Infrastructure
  12. Rule-Based Training
  13. Q-learning based Training
  14. Results and Discussions
  15. Conclusion
  16. ...and 2 more sections

Figures (6)

  • Figure 1: Process Flow for MAPDES P2P Energy Trading Simulator
  • Figure 2: Q-learning Algorithm Convergence
  • Figure 3: Total Electricity Cost: P2P & RE vs Non-P2P & No RE
  • Figure 4: Total Electricity Cost: P2P vs Non-P2P (RE-equipped)
  • Figure 5: Electricity demand of Farm Community from the grid in peak hours (5 pm to 7 pm): P2P vs Non-P2P
  • ...and 1 more figures