Taking soybean oil, corn oil, rapeseed oil and olive oil as research objects, the ternary compound model of soybean oil, corn oil, rapeseed oil and soybean oil, corn oil, olive oil was established. The standard fatty acid fingerprint was established by gas chromatography (GC) and common mode method, and the similarity analysis was carried out at the same time. The change of fatty acid content in different proportion of mixed oil was analyzed by gas chromatography and chemometrics. The results showed that the four standard fatty acid fingerprints established in this experiment could better reflect the commonness and difference of different oils. Palmitic acid (5.88%~10.72%), stearic acid (2.31%~4.57%), oleic acid (15.04%~77.45%), linoleic acid (5.92%~53.67%) and α-linolenic acid (0.61%~6.71%) were selected as the characteristic fatty acids. According to the correlation (R2>0.95) and significance (p<0.01), oleic acid and linoleic acid, α-linolenic acid and oleic acid were used as the representative fatty acids of the two ternary compound models. Through the example verification of the regression prediction model of representing fatty acids, the RSD% of the two ternary compound models was verified to be less than 8% and 10% respectively, which indicated that the regression prediction model has strong guiding significance and practical value. This study reflects the differences of fatty acids in different components of vegetable blended oil, which can provide theoretical basis for quality control and detection of edible oil.