With the discovery of more and more reservoirs in the Lower Triassic Baikouquan Formation sandy conglomerate reservoirs in the slope area of Mahu Depression in recent years, prediction techniques for high-quality sandy conglomerate reservoirs have also attracted the attention of many scholars. This article analyzes the rock casting thin sections, mineral composition analysis, and helium gas porosity measurement results of sandy conglomerate reservoirs in the study area. It is found that the content of plastic debris has a significant impact on the porosity and permeability of the rock. As the content of plastic debris in the rock increases, the porosity and permeability of the rock will decrease. At the same time, sensitive parameter optimization of high-quality reservoirs was carried out using logging data. High quality reservoirs exhibited low natural gamma and high resistivity characteristics on conventional logging curves, and a characteristic curve was constructed to characterize the content of plastic debris. This curve can effectively identify high-quality reservoirs in the study area. In this paper, a nonlinear inversion method using artificial intelligence inversion will be introduced, and we will use this method to predicate the characteristic curve in the study area. When the value of the characteristic curve is less than 0.3, we can have a the area is considered have high-quality reservoir.