The vibrating screen is an important mechanical equipment in coal preparation process and its fault signal shows nonlinearity and non-stationarity characteristics with the influence of uncertain factors such as external noise. However, most of the existing fault diagnosis methods are difficult to achieve effective extraction and accurate diagnosis of uncertain fault features. Therefore, a fault diagnosis method for vibrating screen in coal preparation plant based on feature fusion is proposed in this paper. The cloud model and principal component analysis are used to extract the dynamic-static feature information. Based on the stochastic configuration network algorithm, a diagnosis learning network is established in an autonomous way to ensure the fast convergence performance of the model. Finally, the superiority of the diagnosis method is verified by the data of the vibrating screen in the coal preparation plant.