基于近红外光谱与颜色特征的炼蜜过程表征方法及系统
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANDONG UNIV
- Filing Date
- 2026-04-10
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies lack interpretability of the model's decision-making mechanism during the honey refining process, resulting in incomplete characterization of the honey refining process, insufficient correspondence between detection results and process states, and difficulty in verifying whether the model's prediction logic conforms to the actual physicochemical reaction pathway.
By employing a fusion method of near-infrared spectroscopy and color features, and combining partial least squares regression and extreme gradient boosting models with a hierarchical interpretation framework, we analyze the driving effect of key spectral wavelengths and color features on quality attributes, and construct an interpretable characterization model for the honey refining process.
It achieves high-precision characterization of the honey refining process, ensures consistency between the model prediction mechanism and the actual reaction path, improves the scientific nature and reliability of honey refining process control, and provides reliable process control and quality traceability capabilities.
Smart Images
Figure CN122024898B_ABST