Physics-Based Compressive Sensing Approach to Monitor Turbulent Flow
Yanglong Lu, Yan Wang
Abstract
Turbulent flow is ubiquitous in nature and also has wide industrial applications. The direct observation of turbulent flow can help us to gain the fundamental insights of its physics. However, the experimental measurement of complete velocity fields in real time with high spatial and temporal resolutions is difficult due to the limitations of sensors and sensing techniques, such as sensor accessibility and sampling rate. Recently, a physics-based compressive sensing (PBCS) approach was proposed to reduce the number of sensors and amount of data collection to monitor temperature distributions with a very high compression ratio. In this Paper, the PBCS mechanism for monitoring turbulent flow is formulated. By recovering the boundary conditions in the Reynolds-averaged Navier–Stokes equations from some limited velocity measurements, the complete velocity field is reconstructed from the model. The application of PBCS to the steady-state turbulent flow is demonstrated and validated with experimental data of direct measurements.