A method for quantitative analysis of organophosphorus pesticides based on enzyme kinetics curve multi-feature extraction

By constructing a multi-source heterogeneous kinetic curve benchmark library and a hierarchical navigation small-world graph index structure, the problem of insufficient reliability assessment of enzyme kinetic methods in food safety testing is solved, and the output of test results with high consistency and transparency is achieved.

CN122241186APending Publication Date: 2026-06-19SHANDONG URBAN CONSTR VOCATIONAL COLLEGE +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SHANDONG URBAN CONSTR VOCATIONAL COLLEGE
Filing Date
2026-04-03
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing enzyme kinetic methods for food safety testing suffer from insufficient reliability assessment of predictive results, inability to provide confidence intervals or quantify uncertainty, and failure to effectively calibrate for heterogeneity across different equipment, reagents, and environments, resulting in insufficient reliability and reasonableness of test conclusions.

Method used

A multi-source heterogeneous dynamic curve benchmark library is constructed. Low-dimensional fingerprint embedding vectors are extracted using a dual-path fingerprint encoder. A dedicated uncertainty interval is generated through a hierarchical navigation small world graph index structure and a weighted kernel regression algorithm. Combined with an interpretability index set, the detection results of adaptive confidence boundaries are output.

Benefits of technology

It achieves highly consistent predictions across devices, reagents, and environmental conditions, provides lightweight uncertainty quantification, enhances the transparency and reliability of test results, and supports intuitive judgment and decision guidance for end users.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122241186A_ABST
    Figure CN122241186A_ABST
Patent Text Reader

Abstract

This invention provides a quantitative analysis method for organophosphorus pesticides based on multi-feature extraction from enzyme kinetic curves. The method includes: constructing a multi-source heterogeneous kinetic curve benchmark library containing responses to a wide range of experimental conditions and standard concentrations; generating low-dimensional embedding vectors with cross-domain robustness through a dual-pathway fingerprint encoder; and achieving efficient indexing and similar sample retrieval based on a hierarchical navigation small-world graph. It also utilizes weighted kernel regression combined with historical prediction residuals to automatically generate concentration uncertainty intervals with asymmetric adaptive confidence boundaries, and introduces structured credibility indicators such as environmental matching degree and neighborhood coverage to achieve interpretability and high reliability of the results. This invention significantly improves the accuracy, consistency, and self-calibration capability of organophosphorus pesticide detection results.
Need to check novelty before this filing date? Find Prior Art