Abstract: Vacuum stacking joint pre-compression method is widely used to reinforce soft ground. During the preloading process, a coupled mechanism involving positive and negative pore-water pressures develops within the soil, and the reinforcement performance is jointly controlled by multiple construction parameters, including drainage board spacing, vacuum intensity, vacuum preloading duration, and loading interval time. However, previous studies have mainly focused on single-factor effects and have not adequately captured the coupled influence of multiple parameters on excess pore-water pressure evolution. In this study, a finite element model was established and validated to investigate the effects of drainage board spacing d, vacuum intensity P_v, vacuum preloading duration T_v, interval time T_i and other construction parameters on the excess pore water pressure under coupled positive-negative pressure conditions. The results show that increasing drainage board spacing significantly enhances excess pore water pressure accumulation, particularly when spacing increases from 0.9 m to 1.0 m. Considering both drainage efficiency and economic performance, a spacing of 0.9 m is recommended. With the increase in P_v, the peak of the excess pore water pressure produced in the soft foundation after continuous stacking basically shows a linear and gradual decline, whereas with the increase in T_i, the excess pore water pressure produced in the soft foundation is basically decreased gradually. The excess pore pressure generated in the soft ground shows a decreasing trend, and the change in the excess pore pressure is not obvious if T_i>15 d, suggesting his value as the optimal interval duration. T_v only affects the excess pore water pressure at the initial load stage; drainage board spacing, vacuum intensity, interval time significantly influence the excess pore water pressure evolution during preloading process. This study provides a scientific basis for the design and construction of vacuum-surcharge preloading treatment.