Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
April 18, 2026Applied SciencesOpen Access

A Multi-Parallel Hybrid Neural Network Model for Short-Term Electricity Price Forecasting Under High Market Volatility

View Full Paper
Ask AI
Bookmark
Share

Authors

NRNeringa RadziukynienėGDGabriele DargeAKArturas Klementavičius

Discussion

Loading...

Member takes

Overview

Demonstrates a new framework that improves electricity price forecasting accuracy under market volatility.

Key Points

  • The main goal is to develop a robust forecasting model for predicting day-ahead electricity prices amid high market volatility.
  • Developed a multi-parallel hybrid forecasting framework with seven neural networks.
  • Utilized a hierarchical structure with six parallel base models feeding into a meta-network.
  • Incorporated a calibration stage using probabilistic error distribution for final forecasts.
  • Validated the model using data from the Lithuanian electricity market from 2020 to 2022.
  • Achieved a notable nMAE of 1.57% and a sMAPE of 34.25%.
  • Showed a consistent stacking effect reducing forecasting residuals with additional models.
  • Demonstrated effective bias mitigation during nighttime and early morning hours.

Cite This Study

Radziukynienė et al. (2026) studied this question.

synapsesocial.com/papers/69e3213840886becb6540753https://doi.org/10.3390/app16083865
View Full Paper
Ask AI
Bookmark
Share