Abstract: To accurately predict the PLC effect and its spatiotemporal evolution behavior of Ti-15Mo alloy, and solve the problems of numerous parameters, low calibration accuracy and poor efficiency of the traditional calibration methods for the McCormick (MC) constitutive model, this paper proposed a three-dimensional finite element modeling method integrating the particle swarm optimization back propagation (PSO-BP) neural network intelligent optimization algorithm and the MC constitutive model. First, the MC constitutive model was embedded into ABAQUS software via the UMAT user subroutine, and a uniaxial tensile finite element model of Ti-15Mo alloy was established. Second, the PSO-BP hybrid intelligent optimization algorithm was adopted to efficiently calibrate seven key parameters of the finite element model. With the coefficient of determination(R²), root mean square error (RMSE) and mean absolute percentage error (MAPE) as evaluation indices, the effectiveness and reliability of the PSO-BP algorithm in predicting the stress level(σ_s), average stress drop amplitude (Δσ_a) and stress drop number (N_d) were verified. Finally, the calibrated finite element model was validated by experimental data obtained from a Gleeble 3500 testing machine, and the spatiotemporal evolution law of the PLC effect in Ti-15Mo alloy was systematically analyzed. The results show that the PSO-BP algorithm converges after 43 iterations. The predicted values of σ_s、Δσ_a and N_d are highly consistent with the true values, with all R² values higher than 0.901 6, all MAPE values lower than 10%, and the minimum RMSE of 0.7. Both the simulated and experimental stress-strain curves exhibit typical serrated fluctuations, and the relative errors of the three characteristic quantities are all less than 9.67%. As the true strain increases from 0.01 to 0.04, Δσ_a rises from 9.63 MPa to 11.46 MPa, and the stress drop period increases from 0.76 s to 0.854 s. This phenomenon is mainly attributed to the enhanced pinning effect of precipitated phases on mobile dislocations in the process of strain accumulation. In addition, the propagation of PLC bands presents a mixed continuous and hopping mode, and the band inclination angle stabilizes at approximately 60°.