Abstract: The aerodynamic mechanisms of rotors are complex, with significant coupling of variables. How to accurately estimate rotor aerodynamic force online becomes one of the key problems of achieving safe autonomous flight of rotor-wing flying robots in high-maneuvering. Therefore, a lumped-parameterized modeling method based on blade element momentum (BEM) theory is proposed. Drawing upon the concept of parameter lumping, the blade element integral formula of BEM model is simplified into algebraic form to improve model computational efficiency and reduce the difficulty of parameter identification. And by retaining the induced velocity of the rotor as a state variable, the resulting model avoids the assumption of quasi-static flight condition, enabling it to adapt to high-speed and high-maneuvering flight states. Simulation results based on real-world maneuvering flight data indicate that the resulting model, compared to both the common lumped parameter model and the model based solely on blade element momentum theory, reduces prediction errors of overall rotor aerodynamic force by 20% and 50%, respectively, with significantly increased computational efficiency compared to the model based solely on blade element momentum theory.