To address the complexities, high costs, and iterative challenges between theoretical research and practical deployment in the recovery process of multi-unmanned surface vehicles (USVs), a virtual-real simulation method and system have been designed based on virtual-real data synchronization mechanisms and high-fidelity rendering via Unity 3D technology. This approach enables simultaneous theoretical research and practical debugging, effectively reducing iteration cycles and lowering costs. The system comprises physical USVs, virtual USVs, and a USV virtual testing platform, and the corresponding method is implemented through this system architecture. Physical USVs generate motion state data through their hardware, software systems, and recovery control strategies, while virtual USVs produce corresponding data through simulations. Through real-time synchronization and interaction of virtual-real data, the constructed USV virtual testing platform facilitates high-fidelity state monitoring, intuitive visualization, and high-confidence validation of recovery strategies. Comprehensive experiments indicate that this method can replace large-scale physical deployments with small-scale physical tests, fully integrating the advantages of real hardware and virtual environments to enable efficient, safe, and low-cost validation of multi-USV recoveries. Furthermore, the designed virtual-real simulation method and system provide a convenient and efficient testing platform for validating the recovery process of larger-scale USV formations and clusters.