An individual deviation and group behavior statistical field coupling-based target anomaly detection method, system, device and storage medium

By coupling individual behavior with the statistical field of group behavior, the anomaly detection process is dynamically adjusted, which solves the problems of high false alarm rate and insufficient robustness in the existing technology, and achieves high accuracy and real-time adaptive target anomaly detection.

CN122333282APending Publication Date: 2026-07-03BEIJING ZHURUI TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
BEIJING ZHURUI TECHNOLOGY CO LTD
Filing Date
2026-05-07
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing technologies lack the ability to incorporate the overall state of the group into the individual anomaly detection logic in dynamic environments, resulting in high false alarm rates, insufficient robustness, and poor real-time adaptability. Furthermore, static feature splicing methods lack dynamic mechanisms.

Method used

By acquiring multidimensional behavioral feature data, an individual behavior distribution model and a group behavior statistical field model are established. Individual deviation and group statistical variables are calculated. A modulation function is used to generate coupled anomaly indicators, and the anomaly judgment indicators and sensitivity are dynamically adjusted to achieve real-time detection.

Benefits of technology

It improves the accuracy of anomaly identification, reduces the false alarm rate, enhances the robustness and adaptability of the system, is suitable for real-time deployment on edge devices, and reduces the dependence on cloud computing power.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122333282A_ABST
    Figure CN122333282A_ABST
Patent Text Reader

Abstract

This invention relates to the field of target anomaly detection technology, and discloses a target anomaly detection method, system, device, and storage medium based on the coupling of individual deviation and group behavior statistical field. The method involves acquiring multidimensional behavioral feature data from a continuous time series of the target to be detected, constructing an individual behavior distribution model and a group behavior statistical field model, calculating individual deviation, and generating a coupled anomaly index by combining the group behavior statistical field variables through a modulation function. Anomaly judgment control parameters are generated or updated based on the coupled anomaly index and input into an anomaly judgment function. The processor outputs the anomaly detection result and response control action. The model supports dynamic updates based on sliding windows, event triggering, or historical data. This invention can improve the accuracy of anomaly detection in complex dynamic scenarios and reduce the false alarm rate, and is suitable for edge computing, cloud-edge collaboration, and multi-target real-time monitoring scenarios.
Need to check novelty before this filing date? Find Prior Art