Digital measurement and quality transfer modeling of honey refining process oriented to intelligent manufacturing

To facilitate the intelligent transformation of traditional Chinese medicine (TCM) manufacturing, this study addresses the challenges in the honey refining process—namely, the reliance on manual experience, difficulties in quantifying multi-dimensional quality attributes, and unclear quality transfer mechanisms. This study, using real-world samples from the production of Tongren Niuhuang Qingxin Pills, established a digital measurement system for physical and chemical quality attributes of the refining process and constructed a pilot-scale quality transfer model. For physical characterization, optimized parameters were determined for dual-mode rheological testing (rotation: shear rate 50 s^-1; oscillation: frequency 5 Hz, strain 50%) to measure honey viscosity. For chemical attributes, a high-performance liquid chromatography (HPLC) method was developed to quantify pH, moisture, fructose, glucose, sucrose, maltose, and 5-hydroxymethylfurfural (5-HMF). Based on 984 data points from both atmospheric and vacuum refining processes, a kinetic model of quality transfer was constructed. The dynamic viscosity (η) showed strong correlations with moisture (S) and temperature (T) (R² > 0.998 3), effectively representing the evolution of physical properties. Additionally, the chemical kinetic model indicated that vacuum refining at low temperature, combined with honey sources low in fructose and monosaccharide/disaccharide ratios, can significantly inhibit Maillard reaction rates. This study presents a systematic digital measurement and modeling framework for the honey refining process, providing technical support for process optimization and quality consistency in TCM manufacturing under the paradigm of intelligent production.