Abstract: Tight sandstone reservoirs have the characteristics of strong heterogeneity,poor physical properties,and difficulty in exploration and development.In order to find the high-brittleness section of tight sandstone reservoirs in the Weibei oilfield and fracture this kind of reservoirs,this paper proposes a method based on ANN (artificial neural network) model for shear wave prediction under the condition of lacking suitable brittleness prediction methods for tight sandstone reservoirs in the Weibei oilfield at present.The predicted value is highly consistent with the measured value,and the brittleness index of each well in the study area is calculated by the elastic parameter method further.For the purpose of improving the accuracy of the brittleness index predicted by this method,X-ray diffraction full-rock analysis of fewer wells in the study area is utilized,and it is concluded that quartz and carbonate rocks are the main brittle minerals of the Yanchang formation in the study area."(quartz+carbonate) content/ total minerals " are adopted to calculate the rock brittleness index and then improve the brittleness index predicted by the elastic parameter method.This technique which takes advantage of the balance between the mineral composition method and the elastic parameter method not only improves the prediction accuracy but also makes up for the lack of array acoustic logging and whole rock analysis data.This method was used to predict the brittleness of tight sandstone reservoirs in the WB2 well area of the Yanchang formation in the Weibei oilfield, and high-brittleness sections of WB52 and WB49 were further chosen to be fractured.It is shown that the production stimulation effect was obvious after fracturing,which is of great significance for guiding hydraulic fracturing.The method and process proposed in this paper have strong application and promotion value.