• Corpus ID: 237386357

End-to-End Demand Response Model Identification and Baseline Estimation with Deep Learning

  title={End-to-End Demand Response Model Identification and Baseline Estimation with Deep Learning},
  author={Yuanyuan Shi and Bolun Xu},
  • Yuanyuan Shi, Bolun Xu
  • Published 2 September 2021
  • Computer Science, Engineering, Mathematics
  • ArXiv
This paper proposes a novel end-to-end deep learning framework that simultaneously identifies demand baselines and the incentive-based agent demand response model, from the net demand measurements and incentive signals. This learning framework is modularized as two modules: 1) the decision making process of a demand response participant is represented as a differentiable optimization layer, which takes the incentive signal as input and predicts user’s response; 2) the baseline demand forecast… 

Figures and Tables from this paper


Deep Learning for Household Load Forecasting—A Novel Pooling Deep RNN
A novel pooling-based deep recurrent neural network is proposed in this paper which batches a group of customers’ load profiles into a pool of inputs and could address the over-fitting issue by increasing data diversity and volume.
Examining uncertainty in demand response baseline models and variability in automated responses to dynamic pricing
This paper defines several Demand Response parameters, which characterize changes in electricity use on DR days, and develops a metric to determine how much observed DR parameter variability is attributable to real event-to-event variability versus simply baseline model error.
Probabilistic individual load forecasting using pinball loss guided LSTM
A probabilistic load forecasting method for individual consumers is proposed to handle the variability and uncertainty of future load profiles, using a deep neural network, long short-term memory (LSTM), to model both the long-term and short- term dependencies within the load profiles.
Mechanism design for self-reporting baselines in Demand Response
It is shown that truthful reporting of baseline and marginal utility is both incentive compatible and individually rational for every consumer, which establishes the correct baseline and the aggregator is able to meet any random load reduction requirement reliably.
Demand response in the smart grid: The impact of consumers temporal preferences
This work analyses Demand Response as a game theoretic environment and studies the equilibria of the game between consumers with preferences within two different dynamic pricing mechanisms, respectively the daily proportional mechanism introduced by Mohsenian-Rad et al, and an hourly proportional mechanism.
A Contract Design Approach for Phantom Demand Response
An optimal contract between a demand response aggregator (DRA) and power grid customers for incentive-based demand response is designed and its properties are presented and illustrated through examples.
Exploration of tensor decomposition applied to commercial building baseline estimation
Tensor decomposition of the fan power data identifies components that capture both dominant daily patterns and demand response events, and that are generally more interpretable than those found by principal component analysis.
Statistical Estimation of the Residential Baseline
Demand response on the residential market is becoming a solution to adapt customer consumption to the offer available and therefore lower the electricity peak prices. Tariff incentives and direct
A Sparse Linear Model and Significance Test for Individual Consumption Prediction
This paper proposes a method to improve prediction accuracy of individual users by adaptively exploring sparsity in historical data and leveraging predictive relationship between different users using real-world data from PG&E.
What Do Prosumer Marginal Utility Functions Look Like? Derivation and Analysis
A method is proposed to derive marginal utility functions based on the optimization of a prosumer's energy procurement problem, which reveals that they present a non-linear shape under certain conditions, they cannot be a priori derived, and their form is sensitive to a variety of factors.