A long-time dependence modeling and cross-branch fusion sea temperature prediction method and system

By combining a deep convolutional encoder, two-dimensional wavelet decomposition and wavelet attention module, ConvLSTM and Informer temporal encoder, the problems of long-term dependence and cross-branch fusion in sea surface temperature prediction are solved, and high-precision ocean spatiotemporal sequence prediction is achieved. It is suitable for sea surface temperature prediction, climate trend analysis and ocean dynamic field evolution modeling.

CN121639947BActive Publication Date: 2026-04-28NANJING UNIV OF INFORMATION SCI & TECH
2 Cites 0 Cited by

Patent Information

Application Number
CN202610159340.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-02-04
Publication Date
2026-04-28
Estimated Expiration
2046-02-04

AI Technical Summary

Technical Problem

Existing sea surface temperature prediction technologies have shortcomings in long-term dependency modeling and cross-branch fusion, making it difficult to effectively capture complex temporal dynamic changes and multi-scale spatial details, resulting in fuzzy prediction results and insufficient information interaction.

Method used

A deep convolutional encoder is used for spatial feature extraction. Two-dimensional wavelet decomposition and wavelet attention module are combined to enhance feature interaction. A ConvLSTM module is introduced to strengthen high-frequency dynamic modeling. Long-term dependency modeling is achieved through an Informer temporal encoder. At the same time, multi-scale loss and frequency domain constraints are introduced during the training phase.

Benefits of technology

It significantly improves the multi-scale modeling capability of sea surface temperature prediction and the local texture clarity and temporal consistency of prediction frames, enhances the stability and scalability of the model, and has good generalization ability and engineering feasibility.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The application discloses a kind of long time dependence modeling and cross-branch fusion sea temperature prediction method and system in the technical field of marine weather intelligent prediction and deep learning modeling.The method comprises the following steps: using deep convolutional encoder to encode the spatial features of standardized sea temperature time series image sequence, obtaining multi-level spatial features; the multi-level spatial features are two-dimensionally discrete wavelet decomposed, to obtain low-frequency approximation representation, horizontal singular feature, vertical singular feature and diagonal edge feature, then input into wavelet attention module, to obtain short-term time series enhanced features;Multi-level spatial features and low-frequency approximation representation are input into ConvLSTM module, to obtain high-frequency dynamic enhancement features;Short-term time series enhanced features and high-frequency dynamic enhancement features are input into cross-branch attention fusion module, to obtain fusion features, then input into Informer time series encoder and decoder in turn, to obtain prediction results.
Need to check novelty before this filing date? Find Prior Art