Abstract:

For multivariable and nonlinear modern complex systems, a hybrid modeling strategy is one of the most effective ways to get accurate and reliable models. In those existing hybrid modeling methods, it is difficult to clarify the scopes of prior models and nonparametric models. Besides, there is lack of uniform criterion for the selection of proper linearization methods when analyzing nonparametric models. To solve the above issues, a hybrid modeling method based on physical effect analysis and nonlinearity measure is proposed, which is applied to the quadrotor helicopter. It clarifies the scopes of the prior model and the nonparametric model according to the influnce of various physical effects on the flying attitudes; and uses the fuzzy inference algorithm to select proper linearization methods in accordance with nonlinear degrees of different physical effects. The simulation results show the accuracy and reliability of the proposed hybrid modeling strategy.