In order to judge the running state of the grate cooler more accurately, this paper takes the grate cooler as the research object and proposes a method for identifying the working condition of the grate cooler based on RPROP-RF algorithm. First, the typical working conditions are obtained according to the expert knowledge, and the random forest algorithm is selected to model the typical working conditions, and the typical working condition labels are obtained. Then, RPROP neural network is used to predict the grate pressure, and the grate pressure trend label is obtained through trend identification. Finally, a working condition recognition model is established according to the grate cooler working condition evaluation rules, and five working conditions of the grate cooler are identified. In this paper, the actual operation data of a cement plant is used for simulation, and the simulation results verify the effectiveness of the method in this paper.