In order to improve the reliability of diesel engine operation and reduce maintenance costs, a condition-based maintenance decision model for automotive diesel engines based on Weibull proportional failure rate model was established. First, the Weibull distribution was selected as the best distribution of the fault data using the AIC criteria, and then the PCA was used to weighted and fused the various metal content data in the lubricant to obtain a more comprehensive and effective feature covariate, and then the processed data was passed The maximum likelihood estimation obtains the estimated value of the model parameters; and then integrates the principles of minimum unit maintenance cost and maximum availability to establish a comprehensive maintenance optimization model. Based on this model, the optimal prevention time interval and the corresponding condition-based maintenance decision model are obtained. Finally, an example of the same type of diesel engine is analyzed and verified. The results show that the method can be used for decision-making on condition-based maintenance of diesel engines, reducing maintenance costs, and improving operational reliability and availability.