A design method and system for food flavorings and fragrances

By combining a regulatory knowledge graph with a generative artificial intelligence model, the problems of reliance on human experience and lagging regulatory compliance in the design of food flavorings and fragrances have been solved, enabling efficient and safe generation and verification of flavoring formulations, and improving R&D efficiency and consistency of results.

CN121905335BActive Publication Date: 2026-06-30HANGZHOU WEIOU BIOTECHNOLOGY CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HANGZHOU WEIOU BIOTECHNOLOGY CO LTD
Filing Date
2026-03-25
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Existing methods for designing food flavorings and fragrances rely on human experience, resulting in long development cycles, low screening efficiency, delayed regulatory compliance reviews, difficulty in controlling processing-induced risks in advance, and insufficient adaptability of the generated results to actual application scenarios.

Method used

By constructing a regulatory knowledge graph and combining it with a generative artificial intelligence model, candidate formulations that meet multidimensional constraints are generated through unified modeling of target flavor, food category, base material system and processing parameters. The model is then optimized through small-sample verification and feedback update mechanisms.

Benefits of technology

It significantly reduces invalid candidate solutions, shortens the R&D cycle, improves the consistency of design results and the efficiency of engineering transformation, and enhances the applicability, compliance and safety of flavor and fragrance formulations in actual food production.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121905335B_ABST
    Figure CN121905335B_ABST
Patent Text Reader

Abstract

This invention relates to the field of food technology, and more particularly to a design method and system for food flavorings and fragrances. The method acquires a target flavor description, food category, base system parameters, processing parameters, and design boundary conditions to construct a target condition vector; it then invokes a regulatory knowledge graph and extracts a constraint subgraph; the target condition vector and constraint subgraph are input into a generative model, implementing dynamic constraints during candidate component generation, combination expansion, and content decoding to obtain candidate formulations; finally, sensory matching, compliance margin, processing risk, and stability evaluations are performed on the candidate formulations, and the knowledge graph and generative model are updated based on small-sample validation results to output the final safe formulation. This invention improves design efficiency, compliance, safety, and scenario adaptability.
Need to check novelty before this filing date? Find Prior Art