The invention relates to the technical field of
food quality detection, and discloses a meat product comprehensive scoring method based on
principal component analysis, and the method comprises the steps: firstly obtaining physicochemical index (such as
protein and
fat content) and sensory attribute (such as
color saturation and elasticity index) data of multiple batches of samples, and carrying out the structural arrangement of the two types of data according to batches and index dimensions; respectively carrying out Z-
score standardization, range normalization and other
processing, and extracting a physicochemical index
feature vector and a sensory attribute
feature vector; constructing a dynamic weight strategy through a variance contribution rate and a feature root ratio, and performing weighted fusion on the two types of feature vectors to obtain a fused
feature vector; and finally, inputting the data into a scoring model, and after centesimal
system conversion and distribution calibration, dividing quality grades in combination with an
industry standard threshold to generate a comprehensive scoring result containing
alphanumeric identifiers. The method realizes multi-dimensional data fusion and dynamic empowerment, improves the scoring accuracy, and is suitable for meat product quality evaluation.