Taking the braided river reservoir in the Daqing Oilfield as an example, field outcrop, well logging, and core data were used to characterize the structure of the sand body. By analyzing the contacts between the sand bodies in the braided river reservoir, three connectivity modes, i.e., lateral, longitudinal, and internal connections, were established. Then, a support vector machine (SVM) algorithm was applied to quantitatively predict the connectivity of the sand bodies. The results revealed that by classifying the connectivity evaluation parameters and applying the SVM algorithm, rapid quantitative evaluation of the sand body connectivity was achieved. Verification using dynamic and static data demonstrated that the prediction accuracy of the algorithm reached 88%. Subsequently, a target-based geological modeling method was used to establish a single sand body model based on the level 3rd–4th architecture interface control. Based on the interlayer characterization, the quantitative sand body connectivity results were used as deterministic data. The conductivity between the sand bodies was then assigned to facilitate a refined numerical simulation of the oil reservoir. In summary, in this study, we achieved quantitative characterization and simulation of the connectivity between sand bodies based on coupling of interlayers and conductivities. The numerical simulation results reflect the actual production conditions and provide a technical foundation for subsequent oilfield development and optimization.