抄録
This paper presents an active flow control using Synthetic Jets with fluidic nozzle. The control system is composed of Particle Tracking Velocimetry, Neural Network computing and Pyragas' Delayed Feedback control theory. Synthetic Jets are set on both side of main flow, and connected to fluidic nozzle. The actuators can easily vector a main flow upward or downward by synthesizing suction and blowing actions. This system is applied to water flow with free surface in a rectangular tank having an inlet on the right side wall and outlet at the bottom. Two flow patterns are found to be stable in this tank. Under one condition, inlet flow goes directly to outlet and clockwise circulating flow is observed. Under the other one, while, inlet flow is attached to free surface, resulting in counterclockwise circulating flow. Any other flow pattern is unstable and stabilized to either two stable ones. A target flow pattern is stabilized by the present system although it is naturally unstable.
