Abstract:Day-ahead electricity price forecasting (DAEPF) holds critical significance for stakeholders in energy markets, particularly in areas with large amounts of renewable energy sources (RES) integration. In Japan, the proliferation of RES has led to instances wherein day-ahead electricity prices drop to nearly zero JPY/kWh during peak RES production periods, substantially affecting transactions between electricity retailers and consumers. This paper introduces an innovative DAEPF framework employing a Convolutional Neural Network–Long Short-Term Memory (CNN–LSTM) model designed to predict day-ahead electricity prices in the Kyushu area of Japan. To mitigate the inherent uncertainties associated with neural networks, a novel ensemble learning approach is implemented to bolster the DAEPF model's robustness and prediction accuracy. The CNN–LSTM model is verified to outperform a standalone LSTM model in both prediction accuracy and computation time. Additionally, applying a natural logarithm transformation to the target day-ahead electricity price as a pre-processing technique has proven necessary for higher prediction accuracy. A novel "policy-versus-policy" strategy is proposed to address the prediction problem of the zero prices, halving the computation time of the traditional two-stage method. The efficacy of incorporating a suite of multimodal features: areal day-ahead electricity price, day-ahead system electricity price, areal actual power generation, areal meteorological forecasts, calendar forecasts, alongside the rolling features of areal day-ahead electricity price, as explanatory variables to significantly enhance DAEPF accuracy has been validated. With the full integration of the proposed features, the CNN–LSTM ensemble model achieves its highest accuracy, reaching performance metrics of R2, MAE, and RMSE of 0.787, 1.936 JPY/kWh, and 2.630 JPY/kWh, respectively, during the test range from 1 March 2023 to 31 March 2023, underscoring the advantages of a comprehensive, multi-dimensional approach to DAEPF.

Estimating Input Coefficients for Regional Input–Output Tables Using Deep Learning with Mixup

Estimating high-resolution interregional input–output tables: a Bayesian spatial approach

Nonparametric Probabilistic Forecasting of Regional Photovoltaic Power Based on Spatial Clustering and Combining Quantile Regression

R-Mixup: Riemannian Mixup for Biological Networks

A Regional Industrial Economic Forecasting Model Based on a Deep Convolutional Neural Network and Big Data

Hybrid deep learning models with data fusion approach for electricity load forecasting

Updating China's input-output tables series using MTT method and its comparison

MixupE: Understanding and Improving Mixup from Directional Derivative Perspective

Enhancing Environmental Policy Decisions in Korea and Japan Through AI-Driven Air Pollution Forecast

Machine learning application for estimating electricity demand by municipality

Neural-g: A Deep Learning Framework for Mixing Density Estimation

Enhanced Day-Ahead Electricity Price Forecasting Using a Convolutional Neural Network–Long Short-Term Memory Ensemble Learning Approach with Multimodal Data Integration

Global Mixup: Eliminating Ambiguity with Clustering.

Multi-step power forecasting for regional photovoltaic plants based on ITDE-GAT model

Submeter-level land cover mapping of Japan

Deep learning integration optimization of electric energy load forecasting and market price based on the ANN–LSTM–transformer method

Deep Ensembles Meets Quantile Regression: Uncertainty-aware Imputation for Time Series

A Neural Network System for Transformation of Regional Cuisine Style

IIP-Mixer:Intra-Inter Patch Mixing Architecture for Battery Remaining Useful Life Prediction

Prediction and aggregation of regional PV and wind generation based on neural computation and real measurements

Enhanced Input Modeling for Construction Simulation using Bayesian Deep Neural Networks