An Innovative Genetic Algorithm Approach for Direct Calibration of X-Probe Hot Wires

Abstract

A new genetic algorithm (GA) based method for direct calibration of 2-D hot-wire probe and other instruments that can measure 2-D flow properties is introduced. This new method is an alternative for the previous QR method that is commonly used for calibration of the X-probe hot wires. The proposed GA method resulted in a much smaller error in the velocity estimation while preserving the number of sentences in its calibration equation format. In addition, it preserves the magnitude of its error even when the number of sentences in the calibration equation is decreased while the error in the QR method increases substantially for the same situation. Furthermore, the results were much closer to the real or natural behavior of the hot-wire data when compared with the QR method. Additionally, the GA method was applied for a large range of velocities (10–100 m/s) and the results revealed that the range of error did not vary significantly. The turbulent properties of the flow that were calculated by each method were not the same; however, the results from the GA method were smooth and had reasonable behavior.