A Novel Framework for Significant Wave Height Prediction based on Adaptive Feature Extraction Time-Frequency Network

2025-05-10

AI Summary

To address the accuracy bottleneck in significant wave height (Hs) prediction caused by nonlinear and nonstationary characteristics, as well as test-set data leakage induced by conventional explicit time-frequency decomposition, this paper proposes the Adaptive Feature Extraction Time-Frequency Network (AFE-TFNet). AFE-TFNet introduces a novel two-stage encoder that seamlessly integrates wavelet and Fourier transforms—eliminating explicit pre-decomposition. It further incorporates a Dominant Harmonic Sequence Energy Weighting (DHSEW) mechanism to enable adaptive spatiotemporal-spectral feature fusion. Coupled with an enhanced LSTM decoder and a rolling encoder-decoder architecture, the model ensures end-to-end learning consistency and robust forecasting. Evaluated on real-world measurements from three buoys, AFE-TFNet significantly outperforms baseline models, especially in medium- to long-term prediction. It exhibits strong robustness to varying input window sizes and achieves state-of-the-art performance across all four core evaluation metrics.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Precise forecasting of significant wave height (Hs) is essential for the development and utilization of wave energy. The challenges in predicting Hs arise from its non-linear and non-stationary characteristics. The combination of decomposition preprocessing and machine learning models have demonstrated significant effectiveness in Hs prediction by extracting data features. However, decomposing the unknown data in the test set can lead to data leakage issues. To simultaneously achieve data feature extraction and prevent data leakage, a novel Adaptive Feature Extraction Time-Frequency Network (AFE-TFNet) is proposed to improve prediction accuracy and stability. It is encoder-decoder rolling framework. The encoder consists of two stages: feature extraction and feature fusion. In the feature extraction stage, global and local frequency domain features are extracted by combining Wavelet Transform (WT) and Fourier Transform (FT), and multi-scale frequency analysis is performed using Inception blocks. In the feature fusion stage, time-domain and frequency-domain features are integrated through dominant harmonic sequence energy weighting (DHSEW). The decoder employed an advanced long short-term memory (LSTM) model. Hourly measured wind speed (Ws), dominant wave period (DPD), average wave period (APD) and Hs from three stations are used as the dataset, and the four metrics are employed to evaluate the forecasting performance. Results show that AFE-TFNet significantly outperforms benchmark methods in terms of prediction accuracy. Feature extraction can significantly improve the prediction accuracy. DHSEW has substantially increased the accuracy of medium-term to long-term forecasting. The prediction accuracy of AFE-TFNet does not demonstrate significant variability with changes of rolling time window size. Overall, AFE-TFNet shows strong potential for handling complex signal forecasting.

Problem

• Predicting significant wave height with non-linear, non-stationary challenges
• Preventing data leakage during feature extraction in wave forecasting
• Improving accuracy via adaptive time-frequency feature fusion

Innovation

• Combines Wavelet and Fourier Transforms for feature extraction
• Uses Inception blocks for multi-scale frequency analysis
• Integrates time-frequency features via DHSEW weighting