A multi-modal data fusion method and system for intelligent weak current monitoring system

Through multimodal data fusion and intelligent analysis technology, the problem of limited perception capabilities of traditional weak current monitoring systems in complex environments is solved, and more efficient and accurate monitoring and early warning effects are achieved.

CN119783047AActive Publication Date: 2025-04-08SHANDONG SOMAI INFORMATION TECH CO LTD
8 Cites 4 Cited by

Patent Information

Application Number
CN202510274921.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-03-10
Publication Date
2025-04-08
Estimated Expiration
2045-03-10

AI Technical Summary

Technical Problem

Traditional weak current monitoring systems have limited perception capabilities in complex environments and are susceptible to factors such as environmental noise, lighting changes, equipment failures, etc., resulting in reduced monitoring reliability and accuracy.

Method used

The multimodal data fusion method is adopted, and the data preprocessing and feature fusion is achieved by collecting and summarizing weak current data, using a two-stage denoising method and a cross-modal adaptive attention mechanism, and combining unsupervised reconstruction errors and supervised classification anomaly detection methods to achieve deep fusion and intelligent analysis of multimodal data.

Benefits of technology

It improves the perception ability and monitoring accuracy of the weak current monitoring system, enhances the robustness and security of the system, and achieves more efficient security warning and decision-making support.

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

Abstract

The present invention relates to the field of intelligent weak current data processing technology, specifically a multimodal data fusion method and system for an intelligent weak current monitoring system; specifically: collecting multimodal weak current data; performing two-level denoising to eliminate noise interference and retain edge details, and then generating identification items through standardization and hashing to ensure data validity; using a cross-modal adaptive attention mechanism to dynamically allocate the weights of image and time series features, combining a residual fusion strategy to capture the interaction details between modalities, and realizing efficient complementary fusion of image and sensor time series data; generating a comprehensive anomaly score through a joint decision-making mechanism of deep autoencoder reconstruction error and classification network anomaly probability, and realizing high robustness anomaly detection in combination with threshold judgment. The present invention effectively improves the anomaly detection accuracy and system reliability of weak current monitoring scenarios.
Need to check novelty before this filing date? Find Prior Art