Helicopter Rotor Aerodynamic Reduced-Order Model Based on POD-RBFNN Method

The costly and time-consuming helicopter computational fluid dynamics (CFD) calculations makes it difficult to satisfy the requirements of helicopter engineering applications. An aerodynamic reduced-order model (ROM) based on the proper orthogonal decomposition and the radial basis function neural network (POD-RBFNN) is proposed. The ROM can be used to predict the helicopter rotor airloads. The helicopter rotor airloads training samples required by the ROM come from the helicopter CFD. The POD method is used to extract the characteristics modes of the helicopter rotor airloads. The POD modes coefficients of arbitrary rotor control input can be predicted by the RBFNN. The aerodynamic ROM can quickly predict the helicopter rotor airloads at arbitrary control input within a given range. The calculations results show that the ROM has good prediction accuracy and can satisfy the requirements of helicopter engineering calculations.