A method and system for constructing an association index of multi-dimensional heterogeneous data

By constructing a correlation index for multidimensional heterogeneous data, collecting and structurally encapsulating feature vectors of physical source items, task subjects, and execution spaces, performing semantic alignment and temporal overlap calculation, and dynamically adjusting the task execution sequence, the problem of low task scheduling efficiency and hardware security in existing technologies is solved, achieving efficient resource scheduling and system robustness.

CN121979890BActive Publication Date: 2026-06-09FUJIAN RUISIKE MEDICAL TECHNOLOGY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
FUJIAN RUISIKE MEDICAL TECHNOLOGY CO LTD
Filing Date
2026-04-02
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing task scheduling schemes struggle to match the dynamic output fluctuations of physical source items in real time when dealing with heterogeneous data that has strong time-varying characteristics and complex topological constraints. This results in low system performance, and the interrupt recovery logic can easily cause hardware overload or task failure.

Method used

By constructing a correlation index for multidimensional heterogeneous data, collecting and structurally encapsulating feature vectors of physical source items, task subjects, and execution spaces, performing semantic alignment and temporal overlap calculation, dynamically adjusting the task execution sequence, and combining real-time availability features and a safe load reduction mechanism, dynamic re-indexing is achieved.

Benefits of technology

It effectively solves the problem of low precision in traditional static retrieval mechanisms, ensures deep coupling between task retrieval results and resource supply, improves resource scheduling efficiency, avoids the risk of hardware avalanche, achieves a dynamic balance between execution performance and hardware security, and ensures the robustness and execution continuity of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121979890B_ABST
    Figure CN121979890B_ABST
Patent Text Reader

Abstract

This invention relates to the field of electronic digital data processing technology, specifically to a method and system for constructing a correlation index for multidimensional heterogeneous data. The method includes: acquiring the operating status of physical source devices, the time-varying evolution of the task subject, and execution space topological constraints, and performing structured normalization to obtain a heterogeneous feature vector. The target response window of the task subject is identified and semantically aligned with the device output intensity envelope; a correlation coupling degree score is calculated through temporal overlap. Based on this score, tasks are retrieved from a candidate pool, and combined with load limits, a start time and load quota are automatically allocated to generate an initial execution sequence that satisfies both hardware safety and execution efficiency constraints. When an execution flow interruption is detected, a snapshot of the task state is synchronously stored, the residual evolution rate is retrieved, and real-time availability features are coupled for secondary semantic deduction. Under the premise of ensuring that load limits are not triggered, a dynamically corrected reindexed task sequence is generated.
Need to check novelty before this filing date? Find Prior Art