Abstract:Forecasting incoming water demand is a critical step in efficient reservoir management and revenue optimization in large-scale cascade hydropower stations. It depends on multiple factors, such as weather conditions, grid dispatch, and electricity demand, and, in turn, facilitates a range of downstream decision-making, from natural hazards control and water ecology protection to power generation plans. Current efforts mainly rely on methodologies from statistical machine learning or deep neural networks to model the hydrological patterns from historical time series for inflow forecasting. However, existing models are restricted by short-term temporal dependencies and are prone to error accumulation issues due to the underlying autoregressive architecture. Meanwhile, most recent self-attention time-series models fail to achieve real-time inflow forecasting because of tremendous parameters and computational bottlenecks on learning long time series with fine granularity. We propose a novel framework, called DTODE (Dynamic Transformer Ordinary Differential Equations), for capturing nonlinear and non-stationary evolving patterns inherent in hydrological time series. Specifically, we present the dynamic self-attention mechanism combining transformer and ordinary differential equations that simultaneously captures long-range dependencies of observations from a dynamic system perspective. DTODE exploits a continuum of self-attention layers (instead of discrete counterparts) to learn the dynamics of multivariate time series while paying attention to the co-evolving time-related factors. Besides, our model is flexible in inferring the complex states at any time step, allowing us to forecast inflows at multiple time horizons. Comprehensive evaluations on real-world datasets show that DTODE significantly reduces forecasting errors compared to state-of-the-art inflow prediction systems.

Enhancing hydroelectric inflow prediction in the Brazilian power system: A comparative analysis of machine learning models and hyperparameter optimization for decision support

Daily water supply prediction method based on integrated learning and deep learning

Hourly and Daily Urban Water Demand Predictions Using a Long Short-Term Memory Based Model

River flow forecasting for reservoir management through neural networks

Hypertuned temporal fusion transformer for multi-horizon time series forecasting of dam level in hydroelectric power plants

Linear Models Applied to Monthly Seasonal Streamflow Series Prediction

River Flow Forecasting with Constructive Neural Network

An Ensemble Modeling Approach to Forecast Daily Reservoir Inflow Using Bidirectional Long- and Short-Term Memory (Bi-LSTM), Variational Mode Decomposition (VMD), and Energy Entropy Method

A Machine Learning Approach to Forecasting Hydropower Generation

Power substation load forecasting using interpretable transformer-based temporal fusion neural networks

Water and Electricity Consumption Forecasting at an Educational Institution using Machine Learning models with Metaheuristic Optimization

Application of a data-driven DTSF and benchmark models for the prediction of electricity prices in Brazil: A time-series case

Residential energy consumption forecasting using deep learning models

Fine-tuning inflow prediction models: integrating optimization algorithms and TRMM data for enhanced accuracy

Forecasting the Evolution of Hydropower Generation

An Intra-Day Electricity Price Forecasting Based on a Probabilistic Transformer Neural Network Architecture

Dynamic transformer ODEs for large-scale reservoir inflow forecasting

Long-term inflow forecast using meteorological data based on long short-term memory neural networks

A Data-Driven Framework for Small Hydroelectric Plant Prognosis Using Tsfresh and Machine Learning Survival Models

Hydroelectric Generation Forecasting with Long Short Term Memory (LSTM) Based Deep Learning Model for Turkey

Short-Term Electricity Load Forecasting Using the Temporal Fusion Transformer: Effect of Grid Hierarchies and Data Sources