🤖 AI Summary
To address the low accuracy and poor interpretability of multi-scale seasonal electricity load forecasting, this paper proposes RNN(p), a lightweight recurrent neural network featuring structured p-order delayed feedback. Methodologically, RNN(p) generalizes the linear ARX(p) model into a learnable nonlinear RNN architecture—preserving physical interpretability while enhancing modeling capacity. It integrates delay-embedding-based input design, customized gradient optimization, and a multi-scale seasonal modeling strategy to jointly ensure computational efficiency and controllable model complexity. Evaluated on real-world electricity load data, RNN(p) significantly outperforms baselines including ARIMA and LSTM, reducing MAPE by 18.7%. Moreover, it enables attribution analysis and decision traceability via its interpretable structure. These advances establish a new paradigm for high-reliability forecasting in energy markets and financial technology applications.
📝 Abstract
An elementary Recurrent Neural Network that operates on p time lags, called an RNN(p), is the natural generalisation of a linear autoregressive model ARX(p). It is a powerful forecasting tool for variables displaying inherent seasonal patterns across multiple time scales, as is often observed in energy, economic, and financial time series. The architecture of RNN(p) models, characterised by structured feedbacks across time lags, enables the design of efficient training strategies. We conduct a comparative study of learning algorithms for these models, providing a rigorous analysis of their computational complexity and training performance. We present two applications of RNN(p) models in power consumption forecasting, a key domain within the energy sector where accurate forecasts inform both operational and financial decisions. Experimental results show that RNN(p) models achieve excellent forecasting accuracy while maintaining a high degree of interpretability. These features make them well-suited for decision-making in energy markets and other fintech applications where reliable predictions play a significant economic role.
