A multi-sensor fusion behavior monitoring and early warning system for intelligent network point security

CN122454705APending Publication Date: 2026-07-24RUIXI TECH (BEIJING) CO LTD

View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office: CN · China
Patent Type: Applications(China)
Current Assignee / Owner: RUIXI TECH (BEIJING) CO LTD
Filing Date: 2026-04-02
Publication Date: 2026-07-24

AI Technical Summary

Technical Problem

In existing smart network security systems, data from multiple sensors are not synchronized in time and space at the source of collection, making it difficult to align the data accurately, affecting the fusion accuracy and failing to meet real-time response requirements.

Method used

The HarmonyOS distributed soft bus is used to achieve microsecond-level clock synchronization and global timestamps for multiple sensors. It is combined with a lightweight multimodal fusion neural network for feature-level adaptive fusion. The computing tasks are offloaded to nearby devices through a collaborative computing module, and the fusion weights are adjusted by an environment adaptive module to achieve accurate data alignment and real-time response.

Benefits of technology

It achieves precise spatiotemporal alignment of multi-sensor data, improves the accuracy of behavior recognition, reduces recognition latency, and meets the real-time response requirements of smart network security.

✦ Generated by Eureka AI based on patent content.

Smart Images

Figure CN122454705A_ABST

Patent Text Reader

Abstract

The application provides a multi-sensor fusion behavior monitoring and early warning system for smart network point security, comprising a multi-source acquisition module, a time synchronization module, an edge fusion recognition module, a collaborative computing module and an early warning linkage module. The multi-source acquisition module accesses multiple heterogeneous sensors through a distributed soft bus of the Hongmeng; the time synchronization module performs microsecond-level clock synchronization on each sensor and adds a global timestamp; the edge fusion recognition module is deployed in a Hongmeng intelligent network gateway, and realizes feature-level adaptive fusion and behavior recognition through a lightweight multi-modal fusion neural network and a cross-modal attention mechanism; the collaborative computing module unloads the computing task to the adjacent idle equipment through distributed task scheduling when the load exceeds the threshold; the early warning linkage module triggers the local security device linkage when identifying the abnormality. The application realizes accurate spatio-temporal alignment of multi-source data and low-delay edge fusion recognition, and significantly improves the accuracy and real-time performance of security early warning in complex environments.

Need to check novelty before this filing date? Find Prior Art