[Objective] To address the transition issue during the switching process of an power conversion system from off-grid mode to grid-connected mode, this paper studies a pre-synchronization strategy based on virtual synchronous generator (VSG) control. [Methods] Detailed dynamic mathematical models were established for the off-grid operation mode, the pre-synchronization control mode, and the grid-connected operation mode, respectively. Before the grid-connection transition, a pre-synchronization link was introduced to compensate for the voltage amplitude and phase deviations between the inverter output and the grid bus. After incorporating the pre-synchronization proportional-integral (PI) regulator into the original off-grid model, the overall order of the system’s mathematical model was increased from the original second order to third order. The Routh stability criterion was applied to analyze the system stability, and feasible selection guidelines for the control parameters in the pre-synchronization control loop were provided. Furthermore, the root locus method was adopted to investigate the variation of the closed-loop system’s root locus with key parameters after the addition of pre-synchronization control. The considered parameters included virtual inertia, damping coefficient, and the proportional and integral gains of the pre-synchronization control. By analyzing the intersection points of the root loci with the imaginary axis and the movement trends of the dominant pole pairs, the influences of each parameter on system stability and response speed were systematically summarized. [Results] The research results indicated that during the pre-synchronization control process, the inertia time constant of the VSG had a relatively limited impact on the response speed. In contrast, the damping coefficient, along with the proportional and integral gains of the pre-synchronization PI controller, constituted the key factors determining the quality of the system’s dynamic response. The simulation scenarios systematically covered typical operating conditions under different combinations of virtual inertia, damping coefficients, and pre-synchronization PI parameters. The simulation results not only confirmed that the virtual inertia had a weak influence on response speed, but also quantitatively demonstrated the regulation effects of damping and PI parameters on response time, overshoot, and other performance metrics. [Conclusion] The theoretical models and analytical methods established in this paper provide a theoretical basis and practical guidance for the engineering parameter tuning of smooth grid-connection control strategies for microgrids.