Abstract: In order to study the high-dimensional non-linear relationship between the mechanical temperature of the grain pile and its influence factors, the prediction model was established with a random forest algorithm, whose validity was verified by the actual experimental data. By comparing the real value with the predicted value, the results showed that the random forest model had higher prediction accuracy. In order to further demonstrate the accuracy and reliability of the model, the prediction results of the random forest model were compared with the prediction results of the Support Vector Machine and the BP Neural Network model, indicating that the error of the random forest prediction model was the smallest and the regression fitting was the best. Therefore, the model could be applied to predict the mechanical ventilation temperature of grain piles. On this basis, the quantitative relationship between the rate of adjustable factors and the average temperature change rate of the grain pile was calculated, and the critical points set of average grain pile at low temperature condition was calculated, which provided reference for scientifically judging the ventilation timing and temperature control of grain pile.