A Rolling Horizon Approach for a Bilevel Stochastic Pricing Problem for Demand-Side Management

TL;DR

This paper introduces a stochastic bilevel model for demand-side management in electricity networks, using a rolling horizon algorithm to handle large problem sizes and incorporate renewable energy unpredictability.

Contribution

It presents a novel bilevel model combining demand-side management with renewable energy uncertainty and proposes a rolling horizon solution approach.

Findings

Effective demand-shifting through price incentives.

Robust handling of renewable energy variability.

Scalable algorithm suitable for real-world applications.

Abstract

To guarantee the well-functioning of electricity distribution networks, it is crucial to constantly ensure the demand-supply balance. To do this, one can control the means of production, but also influence the demand: demand-side management becomes more and more popular as the demand keeps increasing and getting more chaotic. In this work, we propose a bilevel model involving an energy supplier and a smart grid operator (SGO): the supplier induces shifts of the load controlled by the SGO by offering time-dependent prices. We assume that the SGO has contracts with consumers and decides their consumption schedule, guaranteeing that the inconvenience induced by the load shifts will not overcome the related financial benefits. Furthermore, we assume that the SGO manages a source of renewable energy (RE), which leads us to consider a stochastic bilevel model, as the generation of RE is by nature highly unpredictable. To cope with the issue of large problem sizes, we design a rolling horizon algorithm that can be applied in a real context.