In upgrading the aircraft gearbox parts automated grouping line, the introduction of new equipment like the zero-point positioning system and alignment table significantly boosts production efficiency. However, it also brings about increased systematic errors in multi-process machining, affecting machining accuracy. To address these issues, an error modeling and accuracy optimization method for the squaring system under uncertainty input was developed. Firstly, an assembly deviation analysis model, considering differences among sub-master discs of the zeropoint positioning system, was constructed and validated using the K-S test. Subsequently, the study analyzed the accuracy loss during the interchangeability of multiple sub-master discs with master discs. A quantitative function for accuracy loss based on differences in output distribution and out of Spec was defined. Through optimization of the structure of alignment table base, accuracy loss was reduced from 11.53% to 2.33%. Simultaneously, the variation in workpiece runout due to inconsistencies between machining and positioning datum was reduced from 0.117 to 0.039. This significantly improved the performance of the zero-point clamping and setting system, providing essential theoretical support for the analysis and optimization of clamping and positioning accuracy in the production line.