With the integration of the global food industry chain and the improvement of people's living standards, the proportion of imported vegetable oil in the daily diet has gradually increased. Olive oil with rich nutritional value has attracted much attention in vegetable oil products. During the cross-border transportation and customs clearance of imported bulk olive oil, due to the influence of factors such as environment, temperature and time, the unsaturated fatty acids in the olive oil transported in storage may be oxidized. The pulp residue in the virgin olive oil may precipitate in The accumulation of multiple warehouse exchanges has resulted in large differences in the quality of vegetable oil in different locations of the olive oil warehouses and in the same warehouse, which has caused great difficulties for the on-site sampling supervision and quality evaluation of olive oil ports. In response to the need for rapid quality evaluation of bulk olive oil on-site, this paper converts the Raman response intensity into a vector space angle value based on the partial least square method, establishes an olive oil quality index analysis and prediction model, and conducts olive oil based on samples from different sampling points. The fast on-site prediction of quality ensures the precise supervision of bulk olive oil in the import and export links. Firstly, the traditional method is used to measure the acid value, peroxide value and linolenic acid value of olive oil heated for different durations at 220, 240 and 260 ℃ respectively. At the same time, a portable Raman spectrometer is used to detect the corresponding oil sample. Mann spectroscopy, preprocessing the spectral data by means of smoothing filtering and derivation, combining the partial least squares method and the angle measurement method to model and analyze the acid value, peroxide value, and linolenic acid of olive oil. The correlation coefficients of the index models established by the method are all above 0.99, and the relative error range of the angle measurement method does not exceed -5.43%. Seven different samples were randomly selected from imported bulk olive oil for verification. The root means square errors of the three models established by the angle measurement method were 0.025 8, 0.222 8, and 17.064 1 respectively, and the relative error range was -4.71%~5.98%. The results show that the model established by the angle measurement method is more accurate, with better predictability and stability. This method can be applied to the on-site rapid quality identification of imported bulk olive oil quality, improve the quality evaluation of the port on-site supervision link, and provide technical guarantee for the comprehensive evaluation of imported and exported bulk olive oil quality.