油电混动飞机结合燃油与电力的优势，将二次能源统一为电能，具备低噪声、低红外信号的隐身性能，适用于预警机等隐身性能要求高的机种。然而，复杂的运行工况和冲击性负载对其电力系统的稳定性和能量管理带来了挑战。为此，本研究针对防御者小型双发预警机，构建“双发电机+固态锂电池”的串联式油电混动电力系统，提出基于虚拟阻抗法的稳定性优化方法，通过小信号建模和阻抗分析建立系统稳定性模型，利用粒子群算法优化虚拟阻抗值提升系统稳定性。并在此基础上，设计了一种基于模型预测控制（MPC）的能量管理策略，以系统稳定性优化和燃油经济性最优为目标，能够在多约束条件下实现实时优化控制。通过对比基于状态机和遗传算法的能量管理策略，所提出的MPC策略能够节约6.33%的燃油，运行速度提高了9倍，在起飞、爬升、巡航等各个阶段，稳定性裕度都得到了提高，改善了系统稳定性和控制性能，同时在PLECS-RT BOX平台中验证了虚拟阻抗算法的可行性。

The hybrid electric propulsion system for hybrid electric aircraft combines the advantages of both fuel and electric power, unifying secondary energy into electrical energy. This system offers low noise and low infrared signature stealth performance, making it suitable for aircraft with high stealth requirements, such as early warning aircraft. However, the complex operating conditions and impulsive loads present challenges to the stability of the power system and energy management. To address this, this study constructs a series hybrid electric power system with a "dual-generator + solid-state lithium battery" configuration for the Defender small twin-engine early warning air-craft. A stability optimization method based on the virtual impedance approach is proposed, where a system stabil-ity model is established through small-signal modelling and impedance analysis. Particle swarm optimization (PSO) is employed to optimize the virtual impedance values, enhancing system stability. Furthermore, a Model Predictive Control (MPC)-based energy management strategy is designed, aiming for both system stability optimization and optimal fuel efficiency. By comparing the proposed MPC-based energy management strategy with state machine and genetic algorithm-based strategies, the MPC approach achieves a 6.33% reduction in fuel consumption and a ninefold increase in computational speed. Moreover, it enhances stability margins across all flight phases, includ-ing take off, climb, and cruise, thereby improving overall system stability and control performance. The feasibility of the virtual impedance algorithm is validated on the PLECS-RT BOX platform.