[Objective] The challenge of achieving global efficiency optimization in a single-input multiple-output wireless power transfer (SIMO-WPT) system under multiple load voltage constraints is addressed. [Methods] A full-link transmission efficiency model, including coil and diode losses, was constructed. The control mechanisms of phase-shifting on the transmitter side and Buck-Boost regulation on the receiver side were analyzed. A genetic-interior point collaborative optimization algorithm was proposed. This algorithm combined the global optimization capability of the genetic algorithm with the rapid convergence of the interior point method, thereby avoiding the issues of local optima and initial point sensitivity. [Results] The simulation and experimental results demonstrated that the proposed genetic-interior point collaborative optimization algorithm achieved faster convergence compared to traditional approaches and strictly converged to the global efficiency optimum. [Conclusion] The efficiency model developed in this study covers the key loss components of WPT systems, making it adaptable to scenarios with varying load numbers, voltage constraints, and coil parameters, thus demonstrating strong generalizability. The proposed genetic-interior point collaborative optimization algorithm effectively addresses the challenge of global efficiency optimization in SIMO-WPT systems under load voltage constraints, providing a feasible reference for the efficiency design of similar WPT systems.