Subtopic Deep Dive
Electricity Price Forecasting Models
Research Guide
What is Electricity Price Forecasting Models?
Electricity price forecasting models predict day-ahead and intraday electricity prices using time series methods like ARIMA and GARCH alongside machine learning to capture volatility spikes and market dynamics.
This subtopic centers on statistical and ML models for short-term price prediction in competitive electricity markets. Key approaches include ARIMA (Contreras et al., 2003, 1482 citations), wavelet-ARIMA hybrids (Conejo et al., 2005, 963 citations), and GARCH (Garcia et al., 2005, 698 citations). Over 10 highly cited papers from 2002-2020 establish foundational techniques, with recent deep learning comparisons (Lago et al., 2018, 585 citations).
Why It Matters
Accurate price forecasts enable optimal bidding strategies for producers and consumers in spot markets, reducing financial risks from volatility spikes. Weron (2006, 727 citations) models loads and prices statistically for market planning, while Hong et al. (2020, 529 citations) review energy forecasting applications in renewables integration. Contreras et al. (2003) demonstrate ARIMA's role in bidding strategy development, directly impacting market efficiency and grid stability.
Key Research Challenges
Capturing Price Volatility Spikes
Electricity prices exhibit sharp spikes due to supply shortages, challenging linear models like ARIMA. Garcia et al. (2005) apply GARCH to model heteroskedasticity but note limitations in extreme events. Recent deep learning (Lago et al., 2018) improves spike prediction yet struggles with rare occurrences.
Incorporating Market Coupling Effects
Interconnected markets introduce cross-border price dependencies not captured by univariate models. Nogales et al. (2002, 855 citations) use time series for single markets, ignoring couplings. Wu et al. (2007, 843 citations) address stochastic commitments but lack explicit coupling forecasts.
Handling Non-Stationary Series
Price series show trends, seasonality, and regime shifts, degrading forecast accuracy. Conejo et al. (2005) decompose via wavelets before ARIMA, improving stationarity. Weron (2006) reviews stylized facts like multiple peaks, requiring hybrid models for robustness.
Essential Papers
ARIMA models to predict next-day electricity prices
Javier Contreras, Rosa Espínola, Francisco J. Nogales et al. · 2003 · IEEE Transactions on Power Systems · 1.5K citations
Price forecasting is becoming increasingly relevant to producers and consumers in the new competitive electric power markets. Both for spot markets and long-term contracts, price forecasts are nece...
Day-Ahead Electricity Price Forecasting Using the Wavelet Transform and ARIMA Models
Antonio J. Conejo, M.A. Plazas, Rosa Espínola et al. · 2005 · IEEE Transactions on Power Systems · 963 citations
This paper proposes a novel technique to forecast day-ahead electricity prices based on the wavelet transform and ARIMA models. The historical and usually ill-behaved price series is decomposed usi...
Forecasting next-day electricity prices by time series models
Francisco J. Nogales, Javier Contreras, Antonio J. Conejo et al. · 2002 · IEEE Transactions on Power Systems · 855 citations
In the framework of competitive electricity markets, power producers and consumers need accurate price forecasting tools. Price forecasts embody crucial information for producers and consumers when...
Stochastic Security-Constrained Unit Commitment
Lei Wu, Mohammad Shahidehpour, Tao Li · 2007 · IEEE Transactions on Power Systems · 843 citations
This paper presents a stochastic model for the long-term solution of security-constrained unit commitment (SCUC). The proposed approach could be used by vertically integrated utilities as well as t...
Tackling Climate Change with Machine Learning
David Rolnick, Priya L. Donti, Lynn H. Kaack et al. · 2022 · OPUS 4 (Zuse Institute Berlin) · 735 citations
Climate change is one of the greatest challenges facing humanity, and we, as machine learning experts, may wonder how we can help. Here we describe how machine learning can be a powerful tool in re...
Modeling and Forecasting Electricity Loads and Prices: A Statistical Approach
Rafał Weron · 2006 · 727 citations
The book is divided into four chapters. The first one introduces the structure of deregulated, competitive electricity markets with the power pools and power exchanges as the basic marketplaces for...
A GARCH Forecasting Model to Predict Day-Ahead Electricity Prices
Reinaldo C. Garcia, Javier Contreras, M. vanAkkeren et al. · 2005 · IEEE Transactions on Power Systems · 698 citations
Price forecasting is becoming increasingly relevant to producers and consumers in the new competitive electric power markets. Both for spot markets and long-term contracts, price forecasts are nece...
Reading Guide
Foundational Papers
Start with Contreras et al. (2003, 1482 citations) for ARIMA basics, then Conejo et al. (2005, 963 citations) for wavelet enhancements, and Nogales et al. (2002, 855 citations) for time series bidding applications.
Recent Advances
Study Lago et al. (2018, 585 citations) for deep learning vs traditional models, Hong et al. (2020, 529 citations) for energy forecasting outlook, and Weron (2006, 727 citations) for persistent statistical foundations.
Core Methods
ARIMA for autoregressive forecasting (Contreras et al., 2003), GARCH for volatility (Garcia et al., 2005), wavelet transforms for decomposition (Conejo et al., 2005), deep neural networks for nonlinear patterns (Lago et al., 2018).
How PapersFlow Helps You Research Electricity Price Forecasting Models
Discover & Search
Research Agent uses searchPapers with query 'ARIMA GARCH electricity price forecasting' to retrieve Contreras et al. (2003), then citationGraph reveals 1482 citing papers including Garcia et al. (2005), and findSimilarPapers expands to wavelet hybrids like Conejo et al. (2005). exaSearch uncovers niche intraday models from 250M+ OpenAlex papers.
Analyze & Verify
Analysis Agent applies readPaperContent to extract ARIMA parameters from Nogales et al. (2002), verifies volatility claims via verifyResponse (CoVe) against Weron (2006), and runs PythonAnalysis with pandas to replicate GARCH fits from Garcia et al. (2005) data, graded by GRADE for statistical significance (p<0.05 on residuals).
Synthesize & Write
Synthesis Agent detects gaps in spike modeling between ARIMA (Contreras et al., 2003) and deep learning (Lago et al., 2018), flags contradictions in volatility persistence, then Writing Agent uses latexEditText for model comparisons, latexSyncCitations for 10+ refs, latexCompile for publication-ready PDF, and exportMermaid for price forecasting workflow diagrams.
Use Cases
"Replicate ARIMA electricity price forecast from Contreras 2003 with Python code"
Research Agent → searchPapers 'Contreras ARIMA electricity' → Analysis Agent → readPaperContent + runPythonAnalysis (pandas ARIMA fit on sample data) → outputs fitted model plot and residuals stats.
"Write LaTeX review comparing ARIMA vs deep learning price models"
Synthesis Agent → gap detection on Contreras (2003) vs Lago (2018) → Writing Agent → latexEditText (draft sections) → latexSyncCitations (10 papers) → latexCompile → outputs compiled PDF with tables.
"Find GitHub repos implementing GARCH for electricity prices"
Research Agent → searchPapers 'Garcia GARCH electricity' → Code Discovery → paperExtractUrls → paperFindGithubRepo → githubRepoInspect → outputs repo links with runnable Jupyter notebooks for forecasting.
Automated Workflows
Deep Research workflow scans 50+ papers via searchPapers on 'day-ahead price ARIMA GARCH', structures report with citationGraph clusters (Contreras-led ARIMA group, 2002-2005). DeepScan's 7-step chain verifies Nogales et al. (2002) claims using CoVe and runPythonAnalysis on price data. Theorizer generates hybrid model hypotheses from Weron (2006) stylized facts and Lago (2018) DL benchmarks.
Frequently Asked Questions
What defines electricity price forecasting models?
Models predict day-ahead/intraday prices using ARIMA, GARCH, and ML to handle volatility (Contreras et al., 2003; Garcia et al., 2005).
What are core methods in this subtopic?
ARIMA for baseline forecasting (Nogales et al., 2002), wavelet decomposition (Conejo et al., 2005), GARCH for volatility (Garcia et al., 2005), and deep learning (Lago et al., 2018).
What are key papers?
Contreras et al. (2003, 1482 citations) on ARIMA, Conejo et al. (2005, 963 citations) on wavelet-ARIMA, Weron (2006, 727 citations) on statistical modeling.
What open problems remain?
Extreme spike prediction, market coupling integration, and hybrid model scalability (Lago et al., 2018; Hong et al., 2020).
Research Energy Load and Power Forecasting with AI
PapersFlow provides specialized AI tools for Engineering researchers. Here are the most relevant for this topic:
AI Literature Review
Automate paper discovery and synthesis across 474M+ papers
Paper Summarizer
Get structured summaries of any paper in seconds
Code & Data Discovery
Find datasets, code repositories, and computational tools
AI Academic Writing
Write research papers with AI assistance and LaTeX support
See how researchers in Engineering use PapersFlow
Field-specific workflows, example queries, and use cases.
Start Researching Electricity Price Forecasting Models with AI
Search 474M+ papers, run AI-powered literature reviews, and write with integrated citations — all in one workspace.
See how PapersFlow works for Engineering researchers
Part of the Energy Load and Power Forecasting Research Guide