Uneven distribution of machining-induced residual stresses (MIRS) caused by dynamic cutting conditions and material internal properties has a significant impact on fatigue resistance, stress corrosion resistance, and accuracy retention of aerospace structural parts. Owing to the challenges in describing residual stress distribution property, this paper proposes a distribution consistency characterization model with high accuracy, as well as a quantitative evaluation strategy in milling titanium alloy. MIRS reduction surrogate model embedded with mechanism knowledge describes the general profile of MIRS in the explicit form. Compared with the surrogate model, the prediction accuracy of MIRS under Bayesian calibration is improved by more than 25.68%. To evaluate the MIRS distribution consistency, self-consistent indexes are innovatively proposed based on the extraction of four fundamental MIRS features in parametric form. The sensitivity analyses between three machining parameters and self-consistent indexes are carried out, guiding in selecting machining parameters in consideration of MIRS distribution consistency. Consequently, this work can provide a novel insight into determining the optimal machining parameters with comprehensive consideration of MIRS features magnitudes and distribution consistency. According to the proposed procedure, there exists a potential for future extensions to other materials and cutting processes with more complex stress distributions.