Abstract: After the impoundment of the Lize Navigation Hub, the dry-season water level at Wusheng Hydrometric Station was raised by 2.15 m, with the cross-sectional average velocity dropping to as low as 0.11 m/s. The natural stage-discharge single-valued relationship became unstable, and the measurement frequency of the traditional rotor current meter method nearly doubled, significantly increasing the fieldwork intensity. To reduce the flow measurement workload and improve accuracy under low-velocity conditions in the reservoir area, this study utilized Horizontal Acoustic Doppler Current Profiler (H-ADCP) monitoring data to systematically compare the applicability of the traditional correlation-coefficient-based index-velocity optimization method and the proposed Mixed-Poly All-Set Enumeration (MPAE) method for cross-sectional discharge inversion. Through 63 sets of field comparative measurement experiments, the differences between the two methods in index velocity selection and discharge fitting formula establishment were thoroughly analyzed. The results indicate that: (1) three traditional correlation coefficient indices are prone to falling into local optima, where a high correlation coefficient does not necessarily correspond to a small fitting error; (2) the MPAE method employs minimum systematic error and minimum random uncertainty as optimization criteria, identifies the global optimal solution through an all-set enumeration strategy, and the selected fitting formula improves accuracy indicators by more than 2%, significantly outperforming traditional methods. This study confirms the superiority of the MPAE method in multi-index velocity optimization and discharge formula fitting, providing a new technical pathway for automated flow measurement at hydrometric stations in reservoir areas under variable backwater conditions.