Abstract: With the widespread application of new energy and digital technologies in the power system, the randomness, volatility, and decentralization of the system have significantly increased, which bring great challenges to the frequency security and stability of the power grid. In this paper, the impact of DoS attacks on load frequency instability of hybrid photovoltaic and thermal power systems is addressed. Firstly, a network power load frequency control system model is designed, a model predictive control method with buffer is proposed. Secondly, the energy-constrained situation of the DoS attacker is considered, and a denial of service attack model with constraints is designed. Thirdly, through the designed optimal objective cost function, augmented matrices are constructed and classification discussion method based on the DoS attack mode are utilized to perform asymptotic stability analysis of thermal-photovoltaic power system. Finally, the simulation results show that the proposed model predictive control can effectively reduce the impact of DoS attacks on abnormal fluctuations in system load frequency compared with previous load frequency control methods.