Surface Pressure Contour Prediction Using a GRNN Algorithm

Abstract

A new approach based on a Generalized Regression Neural Network (GRNN) has been proposed to predict the planform surface pressure field on a wing-tail combination in low subsonic flow. Extensive wind tunnel results were used for training the network and verification of the values predicted by this approach. GRNN has been trained by the aforementioned experimental data and subsequently was used as a prediction tool to determine the surface pressure. Most of the previous applications of the GRNN in prediction problems were restricted to single or limited outputs, while in the present method the entire planform surface pressure was predicted at once. This highly decreases the calculation time while preserving a remarkable degree of accuracy. The wind tunnel results verify the accuracy of the data offered by the GRNN, which indicates that the present prediction and optimization tool provides sufficient accuracy with modest amount of experimental data.