Consideration of High Proportion Renewable Energy Penetration in Electricity Spot Market Price Spike Probability Forecasting and Risk Early Warning

Abstract

This study addresses the critical challenge of price spike forecasting in electricity spot markets by investigating the influence of high renewable energy penetration on market volatility and the cost-sensitive production processes of energy-intensive industries. With the increasing deployment of smart grids, wide-area sensing systems, and electromagnetic communication infrastructures, accurate prediction of market dynamics has become essential for secure energy management and real-time decision support. A hybrid quantile regression forest–long short-term memory (QRF-LSTM) framework is proposed to estimate price spike probabilities by explicitly incorporating renewable uncertainty factors. The model combines the statistical distribution learning capability of quantile regression forests with the temporal dependency modeling of LSTM networks to capture nonlinear market behaviors and extreme events. Using empirical data from the German electricity market during 2019–2023, the proposed approach achieves a 14.2% improvement in prediction accuracy over the best-performing standalone QRF model under renewable penetration levels exceeding 40%, as evaluated by CRPS metrics. Furthermore, a dynamic risk early warning mechanism adaptively adjusts threshold values according to real-time renewable forecasts and market conditions. The proposed framework provides practical decision support for electricity market participants while offering a scalable methodology for intelligent energy forecasting and risk management in interconnected energy systems supported by electromagnetic information transmission and smart communication networks. The results further reveal that price spike probabilities increase nonlinearly once renewable penetration exceeds critical thresholds, particularly when wind power variability during evening ramping periods introduces significant operational uncertainty.