Real Time Data Monitoring of Wind Tunnel Calibration and Its Extention, Based on General Regression Neural Network

Wind tunnel tests with good flow conditions have great roles in the design and optimization of models of flying vehicles. It is crucial to obtain test section flow parameters, such as velocity and pressure distribution, as well as uniformity along with the direction of the flow in the calibration tests. This process is very expensive and time consuming due to large Mach number range, especially for relatively large sizes of wind tunnel test section. To overcome these problems, the calibration tests were done in the specified sections and Mach number and the results were extracted for other sections and Mach numbers by soft computing. With real time monitoring of wind tunnel data, based on GRNN and the network training, the related data in other sections and Mach numbers were estimated. It was shown that the GRNN is more capable with respect to classical methods for estimating non-linear functions, such as least square method. Cost, time, and wind tunnel calibration tests were decreased extensively using this intelligent technique.