The invention discloses a
grease quantitative
analysis method based on a near-
infrared transmission and
reflection spectrum technology, which comprises the following steps: S1, collecting a pluralityof
grease samples, and carrying out calibration wet chemical detection; s2, formulating a
grease fatty acid reference standard of each grease index; s3, synchronously carrying out near
infrared spectrum scanning on the collected grease sample; s4, through calibration
software OPUS8.1, according to the wet chemical detection value calibrated by the grease sample in the S1 and the near
infrared spectrum in the S3, carrying out
spectrogram preprocessing; according to a spectral residual error, a
Mahalanobis distance and a component density value, rejecting an abnormal spectrum and a wet chemicalvalue, establishing a grease
quantitative model by applying a partial
least squares method; s5, performing
algorithm verification on the established
quantitative model, and determining an optimal
quantitative model; s6, performing blind
verification on the quantitative model; s7, performing near infrared spectrum scanning on the grease sample to be detected to obtain a prediction result; and S8, analyzing the
fatty acid result of the to-be-detected grease sample obtained by the quantitative model according to the grease
fatty acid reference standard formulated in the S2, and identifying the adulteration condition.