Since the reservoir described by physical characteristics directly affects the reservoir development plan, related research and dynamic management, etc., the staff can deduce various parameters of the reservoir through seismic data and early warning data. However, under the influence of cement types and other factors, porosity and permeability are difficult to accurately detect and predict, and the relationship between them cannot be determined. With the rapid development of modern science and technology, some scholars proposed to use manual skills to build a reservoir storage geological parameter analysis and prediction system to improve paid description and quantitative description of residual, so as to improve the accuracy of prediction analysis. After understanding the CMAC neural network model, according to the current situation of reservoir engineering sample analysis and management in China in recent years, this paper mainly explores the batch processing CMAC neural network model based on global dynamic information and its application effect, in order to provide technical support for reservoir engineering construction management in the new era.