In response to the problem of poor lubrication caused by the deterioration of lubricating oil in mechanical equipment, based on the monitoring of the electrochemical impedance characteristics of lubricating oil, a partial least squares regression analysis method is proposed to process the oil impedance monitoring data, and a grey equal dimensional filling model is used for deterioration prediction. Taking a certain type of natural gas compressor lubricating oil as the research object, a PLS model was constructed and data fitting analysis was carried out by initializing the electrochemical impedance data at multiple frequency points and extracting principal components. The results showed that the linear regression variance fitted by the partial least squares regression analysis method could accurately predict the deterioration state of the natural gas compressor lubricating oil. The grey equal dimensional filling model was used to predict the data, and combined with the PLS model for diagnostic analysis, the deterioration trend of the lubricating oil in the future period could be obtained, providing a basis for planning oil change cycles and formulating mechanical equipment maintenance strategies.