Analysis of Hot-Wire Anemometer Turbulent Signals by Means of Delay Based Networks

A Neural Network based system for reconstructing turbulent signals obtained by a Hot Wire Anemometer in the shear layer surrounding a turbulent free yet of air is presented. The flow-field used as work bench presents very complex turbulence dynamics making the measured signals very difficult to predict. The results presented here show a very good agreement in prediction and present a high potential as an analysis tool.