Abstract: [Objective]To establish an accurate and rapid detection method for sensory odor of raw milk and provide technical support for its quality control.[Method]Due to the wide application advantages of electronic nose technology in the fields of dairy adulteration detection,component analysis,and odor research on quality changes during processing and storage,this study adopted a comprehensive analysis strategy combining electronic nose technology with traditional instrumental analysis and sensory evaluation to carry out raw milk quality detection experiments.[Result]Electronic nose technology showed significant advantages in raw milk quality detection：it could quickly identify abnormal samples,effectively avoid errors caused by subjective human factors,and greatly improve the efficiency of practical detection operations.The detection data of 10,230 raw milk samples from pastures from March to December 2023 showed that the overall compliance rate reached 91.6%,and the non-compliant samples detected by electronic nose were verified by artificial sensory evaluation with a compliance rate of 100%.[Conclusion]A spectral model of normal raw milk was constructed through multiple experiments.By comparing with the spectral model of abnormal samples,the characteristic volatile compounds in abnormal raw milk（such as 3-ethyl-2-methylpyrazine in sour samples and ethanol in bean-flavored samples）were identified,providing a scientific and reliable detection basis for the sensory quality evaluation of raw milk.

Key words: electronic nose, raw milk, component detection, sensory evaluation, volatile compounds