A message early warning system and method for quantitative risk control

By training an anomaly link prediction model and resource load profile, the persistent services of edge computing nodes are dynamically controlled, solving the problems of data loss and improper resource allocation in high-frequency trading and achieving efficient operation of quantitative risk control.

CN121814610BActive Publication Date: 2026-05-26ACCELECOM INFORMATION & TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
ACCELECOM INFORMATION & TECH CO LTD
Filing Date
2026-03-09
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

In high-frequency trading scenarios, edge computing nodes are susceptible to external environmental interference, which can lead to data loss. Their resource load cannot support the full range of business processing, and their resource configuration lacks dynamic adaptation, which affects the efficiency of quantitative risk control.

Method used

By acquiring network link characteristic data and resource load data of edge computing nodes, an abnormal link prediction model is trained. Combined with resource load profiles, persistent services are dynamically controlled, anomalies are predicted in real time, and pre-persistence is initiated to optimize resource allocation.

Benefits of technology

It enables proactive prevention of data loss before anomalies occur, ensures the integrity of data sources for risk control judgments, avoids resource shortages or redundant waste, and guarantees the efficient advancement of quantitative risk control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121814610B_ABST
    Figure CN121814610B_ABST
Patent Text Reader

Abstract

This invention discloses a message early warning system and method for quantitative risk control, relating to the field of data transmission technology. The invention collects network link characteristics and raw resource load data of nodes, preprocesses and trains a dedicated abnormal link prediction model, combines real-time data to predict abnormal time points and initiate pre-persistent services, transforming traditional passive persistence into proactive prevention and avoiding the risk of core risk control data loss due to insufficient node load. Simultaneously, it retrieves historical load data based on the node's unique identifier to construct a resource load profile, dynamically controlling service start-up and shutdown through real-time load and baseline range verification, establishing a precise load-service matching mechanism. Furthermore, it calculates the node's early warning base value and the number of comprehensive early warnings, dynamically adjusts service terminal resource allocation based on early warning thresholds, and configures a pre-persistent data recycling timer to automatically clean up data, solving the resource supply imbalance problem, saving node storage resources, and comprehensively ensuring the efficient advancement of quantitative risk control.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Risk control early warning method, system and equipment, and computer readable storage medium

    CN113657710A

  • IT asset fault propagation prediction method and system based on dynamic evolution of knowledge graph

    CN120821591A