Effective Self-Attention-Based Deep Learning Model with Evolutionary Grid Search for Robust Wave Farm Energy Forecasting

2025-07-13

AI Summary

To address the critical challenge of low prediction accuracy for wave energy converter (WEC) power generation—thereby compromising grid integration stability—this paper proposes a self-attention-enhanced Convolutional Bidirectional LSTM (SA-CBiLSTM) model. The architecture jointly exploits spatial features of WEC arrays and temporal dynamics of wave conditions, while incorporating an evolutionary grid search for automated hyperparameter optimization. Evaluated on real-world measurement data from four Australian sites—Adelaide, Sydney, Perth, and Tasmania—the model achieves R² scores of 91.7%, 91.0%, 88.0%, and 82.8%, respectively, significantly outperforming ten state-of-the-art machine learning and deep learning baselines. This work presents the first integration of self-attention with convolutional BiLSTM for wave energy forecasting, effectively balancing local pattern extraction and long-range dependency modeling. The framework demonstrates cross-regional robustness and engineering scalability, offering reliable predictive support for high-penetration renewable energy grid integration.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Intelligent Robots: State Estimation
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Achieving carbon neutrality, a key focus of UN SDG #13, drives the exploration of wave energy, a renewable resource with the potential to generate 30,000 TWh of clean electricity annually, surpassing global demand. However, wave energy remains underdeveloped due to technical and economic challenges, particularly in forecasting wave farm power output, which is vital for grid stability and commercial viability. This study proposes a novel predictive framework to enhance wave energy integration into power grids. It introduces a hybrid sequential learning model combining Self-Attention-enhanced Convolutional Bi-LSTM with hyperparameter optimization. The model leverages spatial data from Wave Energy Converters (WECs) and is validated using datasets from wave farms in Adelaide, Sydney, Perth, and Tasmania, Australia. Benchmarked against ten machine learning algorithms, the model achieves superior accuracy, with R2 scores of 91.7% (Adelaide), 88.0% (Perth), 82.8% (Tasmania), and 91.0% (Sydney). It outperforms conventional ML and deep learning methods, offering robust and scalable predictions for wave energy output across diverse marine environments, supporting reliable integration into energy systems.

Problem

• Enhancing wave farm energy forecasting accuracy
• Overcoming technical challenges in wave energy prediction
• Integrating wave energy reliably into power grids

Innovation

• Self-Attention-enhanced Convolutional Bi-LSTM model
• Evolutionary grid search for hyperparameter optimization
• Spatial data from Wave Energy Converters