Fuzzy logic algorithm for flush air data sensing system

To overcome the disadvantages of traditional air data sensing system, the embedded flush air data sensing (FADS) system is investigated. Based on wind tunnel experiments and computational fluid dynamics(CFD) simulation results of an all-wing aircraft, the fuzzy logic modeling algorithm for the FADS system is studied. By using pressure or pressure coefficients on the model surface as input data while angle of attack, angle of sideslip, inlet velocity and altitude as output data, a fuzzy logic structure is built for the FADS system with adaptive and fixed shape parameters membership functions as the fuzzy model's component and parameters of the system solved by a hybrid learning procedure (gradient descent and least squares estimate method). The modeling results show that, compared with the methods with fixed shape parameters membership functions and inner functions parameters solved by gradient descent, the method with adaptive membership functions has higher modeling accuracy and efficiency.