Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Simplifying Preference Elicitation in Local Energy Markets: Combinatorial Clock Exchange

Shobhit Singhal, Lesia Mitridati

TL;DR

This work tackles the challenge of engaging prosumers with complex, interdependent preferences in local energy markets by proposing a multi-product combinatorial exchange (CCE) cleared with linear pricing. It blends an iterative price-discovery mechanism with machine learning (MLCCE) to efficiently learn prosumer value functions via monotone MVNNs and accelerate convergence, while preserving transparency and privacy through package queries. Theoretical insights (duality-gap bounds for non-concave utilities) are complemented by numerical experiments showing that linear prices approximately clear in large markets, that a joint multi-product market outperforms product-specific markets, and that ML-based updates reduce convergence time. The results indicate that the proposed framework can deliver welfare gains, scalability, and practical participation benefits for DER-rich grids, with a structured approach to balancing trade and accounting for externalities.

Abstract

As distributed energy resources (DERs) proliferate, future power system will need new market platforms enabling prosumers to trade various electricity and grid-support products. However, prosumers often exhibit complex, product interdependent preferences and face limited cognitive and computational resources, hindering engagement with complex market structures and bid formats. We address this challenge by introducing a multi-product market that allows prosumers to express complex preferences through an intuitive format, by fusing combinatorial clock exchange and machine learning (ML) techniques. The iterative mechanism only requires prosumers to report their preferred package of products at posted prices, eliminating the need for forecasting product prices or adhering to complex bid formats, while the ML-aided price discovery speeds up convergence. The linear pricing rule further enhances transparency and interpretability. Finally, numerical simulations demonstrate convergence to clearing prices in approximately 15 clock iterations.

Simplifying Preference Elicitation in Local Energy Markets: Combinatorial Clock Exchange

TL;DR

This work tackles the challenge of engaging prosumers with complex, interdependent preferences in local energy markets by proposing a multi-product combinatorial exchange (CCE) cleared with linear pricing. It blends an iterative price-discovery mechanism with machine learning (MLCCE) to efficiently learn prosumer value functions via monotone MVNNs and accelerate convergence, while preserving transparency and privacy through package queries. Theoretical insights (duality-gap bounds for non-concave utilities) are complemented by numerical experiments showing that linear prices approximately clear in large markets, that a joint multi-product market outperforms product-specific markets, and that ML-based updates reduce convergence time. The results indicate that the proposed framework can deliver welfare gains, scalability, and practical participation benefits for DER-rich grids, with a structured approach to balancing trade and accounting for externalities.

Abstract

As distributed energy resources (DERs) proliferate, future power system will need new market platforms enabling prosumers to trade various electricity and grid-support products. However, prosumers often exhibit complex, product interdependent preferences and face limited cognitive and computational resources, hindering engagement with complex market structures and bid formats. We address this challenge by introducing a multi-product market that allows prosumers to express complex preferences through an intuitive format, by fusing combinatorial clock exchange and machine learning (ML) techniques. The iterative mechanism only requires prosumers to report their preferred package of products at posted prices, eliminating the need for forecasting product prices or adhering to complex bid formats, while the ML-aided price discovery speeds up convergence. The linear pricing rule further enhances transparency and interpretability. Finally, numerical simulations demonstrate convergence to clearing prices in approximately 15 clock iterations.

Paper Structure

This paper contains 22 sections, 1 theorem, 13 equations, 4 figures, 2 algorithms.

Table of Contents

  1. Introduction
  2. Proposed LEM framework: A multi-product combinatorial exchange
  3. Motivation: LEM with combinatorial preferences
  4. Motivating example 1: LEM with joint energy-flexibility products
  5. Motivating example 2: LEM with socio-ecological product differentiation
  6. Combinatorial exchange market clearing
  7. Linear pricing rule and Walrasian equilibrium
  8. Duality gap for non-concave prosumers
  9. Intuitive and scalable preference elicitation
  10. CCE
  11. MLCCE
  12. Value function estimation
  13. Price update
  14. Feasible allocation
  15. Desirable mechanism properties
  16. ...and 7 more sections

Key Result

Theorem 1

For optimization problem eq:marketmodel, where the objective function is a sum of non-convex functions subject to linear equality and inequality constraints, the duality gap is less than or equal to the largest non-convexity udell2016bounding, i.e.,

Figures (4)

  • Figure 1: A schematic of the resulting multi-product trade (incoming arrows indicate consumption and outgoing production) among prosumers in a energy-flexibility market defined in \ref{['sec:resjointmar']} that offers energy and flexibility products across two time slots.
  • Figure 2: Change in social welfare of sequential relative to the joint market as a function of flexibility price forecast errors.
  • Figure 3: Duality gap and imbalance as a function of market size (number of prosumers) for a market with 24 hourly energy products.
  • Figure 4: Imbalance as a function of clock iterations for CCE and MLCCE, in a market with 24 hourly energy products and 120 prosumers (top), and 6 hourly products with 40 prosumers (bottom).

Theorems & Definitions (3)

  • Definition 1: Convex envelope of a function
  • Definition 2: Non-convexity of a function
  • Theorem 1