To address the evaluation difficulty of hydrocarbon saturation in low resistivity reservoirs, an innovative method for calculating oil saturation is proposed using a nuclear magnetic resonance (NMR)-constrained triple-water resistivity model. This model explicitly distinguishes three conductive water phases: movable water, capillary-bound water, and clay-bound water. Resistivity response equations are established for both water-saturated and hydrocarbon-bearing rocks. A cooperative inversion framework is proposed based on NMR and conventional logging data, incorporating high-precision NMR-derived parameters as constraints during conventional logging inversion. The pore component volumes are obtained with the NMR T2 spectrum decomposition and served as a priori information for the nonlinear optimization of porosity exponents. This enables the construction of a pore component inversion algorithm using conventional logging data, thereby extending water saturation calculation applicability in complex reservoirs. The method incorporates data-driven optimization to effectively reduce the reliance on core-based calibration data (mercury injection, petrophysical experiments, cation exchange capacity (CEC) tests). Application in a Bohai Bay Basin low resistivity reservoir demonstrates superior saturation calculation accuracy compared to traditional models. The integration of multi-physics logging inversion with nonlinear optimization effectively enhances conductivity mechanism characterization in reservoirs with complex pore systems, providing a robust technical solution for quantitative evaluation of low resistivity oil reservoirs.