An improved fuzzy logic unsteady aerodynamics modeling method

We aim at the severe nonlinearity, coupling and hysteresis of the aerodynamic forces and moments when an airplane has post-stall maneuver. The modeling method is based on large-amplitude oscillatory wind tunnel test data analysis results and considers various factors that influence the aerodynamic forces and moments. Then we use the quadratic membership function to transform the angle of attack and its first and second-order derivatives, yaw angle and reduced frequency into fuzzy inputs. The aerodynamic forces and moments calculated with fuzzy rules are used as outputs. Compared with other modeling methods, our modeling method can identify the structure and fuzzy logic model in a unitary way and is versatile. The precision of the model satisfies the requirements for given correlation coefficients. We design the time-scale dynamic and inverse control law according to the singular perturbation theory. We also use the Daisy-chain method to allocate the actuator's deflection in an over-actuated system. The comparison of simulation results on the “Cobra” typical maneuver under normal and unsteady aerodynamic forces shows the necessity of establishing a precise unsteady aerodynamic model when an airplane has post-stall maneuver and verifies the good control effect of the dynamic inversion method on a nonlinear system.