The speed of Probabilistic Power Flow (PPF) analysis for hybrid AC/DC grids (ACDCPPF) can be significantly decreased if the corresponding adjustment strategy of Voltage Source Converter (VSC) on the reactive power is not properly designed. To address this issue, a new strategy on the reactive power adjustment of VSC is proposed. A Monte Carlo Simulation (MCS) method is used to model uncertainties in the stochastic behaviors of Photovoltaic (PV) stations and loads. The feasibility and the effectiveness of the improved ACDCPPF are validated using a modified IEEE 9-Bus test system. The simulation results indicate that with the proposed adjustment strategy and the empirical optimal value, the number of AC/DC alternate iterations in a VSC's reactive power limited AC/DC system can be greatly reduced in each sampled Deterministic Power Flow (DPF), which leads to a significant improvement in the computing speed of the ACDCPPF analysis.