The aeroacoustic inverse problem was solved according to the aeroacoustic noises radiated by the rotor. The blade surface loading was given and acoustic analogy method was used to obtain the observation data. A large set of ill-posed equations were solved by singular value decomposition combined with the regularization method. In order to overcome the obvious numerical oscillations
a polynomial function was used to represent the unknown
at the same time the Kutta condition was introduced as a priori information. The results showed that under ideal condition
the pressure of the rotor surface obtained by this method was closer to the input value
especially in the tip region. Under certain interference noise
the maximum relative error of the pressure peak position obtained by this method was reduced from 15% to 4%.