Fuzzy 알고리즘을 이용한 PTV측정법의 개발에 관한 연구

Abstract

PTV can not only probe small scaled eddy structures but also provide spatial information of velocity vectors over the whole flow field. For these reasons, PTV(Particle Tracking Velocimetry) is now widely used in the field of fluid engineering. However, some erroneous vectors are obtained during PTV calculation for the algorithms adopted in the PTV techniques. The main source of the erroneous vectors is the paring method. To obtain the velocity vector distribution, correct particle paring should be made in PTV. Since there are numerous particle images on the two images used for PTV calculation, it is not so easy to identify the same particle between the two images.

In this paper, a new idea based on Fuzzy Theory is adopted to PTV so that the number of correct pairs of particles becomes as much as higher that those of GA-PTV. Mamdani method which is most widely used is used for the calculation of Fuzzy-PTV.

It was verified that the velocity vectors obtained by the constructed Fuzzy-PTV belonged to the original vector data used for the test, which implies that the constructed algorithm is available for the measurement of fluid flows.