Abstract:Drawing inspiration from the lateral lines of fish, the inference of flow characteristics via surface-based data has drawn considerable attention. The current approaches often rely on analytical methods tailored exclusively for potential flows or utilize black-box machine learning algorithms to estimate a specific set of flow parameters. In contrast to a black-box machine learning approach, we demonstrate that it is possible to identify certain modes of fluid flow and then reconstruct the entire flow field from these modes. We use dynamic mode decomposition to parametrize complex, dynamic features across the entire flow field. We then leverage deep neural networks to infer the dynamic modes of the pressure and velocity fields within a large, unsteady flow domain, employing solely a time series of pressure measurements collected on the surface of an immersed obstacle. Our methodology is successfully demonstrated to diverse fluid–structure interaction scenarios, including cases with both free oscillations in the wake of a cylinder and forced oscillations of tandem cylinders, demonstrating its versatility and robustness.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to reconstruct the flow field from the local pressure measurement data on the surface of an object immersed in a fluid. Specifically, the researchers hope to infer the characteristics of the surrounding flow field only through the pressure measurement points on the object's surface. The solution to this problem is of great significance for understanding and predicting fluid - structure interactions, especially in applications involving how marine organisms such as fish use non - visual information to perceive their surroundings. In addition, this technology also has potential value in engineering applications, such as in the fields of aircraft design, ship design, submersibles, offshore structures and heat exchangers, all of which require a deep understanding of structural integrity and fluid dynamics.
The method proposed in the paper combines dynamic system tools and machine - learning techniques, in particular, using Dynamic Mode Decomposition (DMD) and Deep Neural Networks (DNN). The specific steps are as follows:
1. **Dynamic Mode Decomposition (DMD)**: First, through DMD analysis of the pressure measurement data on the object's surface, the main modes of the flow field are extracted. DMD is a technique for approximating unsteady flow. It models the flow field by superimposing linear modes, thereby reducing the time dimension, and these modes usually have physical significance.
2. **Deep Neural Networks (DNN)**: Then, a shallow neural network is used to correlate the modes of the surface pressure field with the modes of the flow field through supervised learning. In this way, even with only the surface pressure measurement data, the entire flow field can be reconstructed by identifying these modes.
3. **Flow Field Reconstruction**: Finally, the identified modes are used to reconstruct the complete flow field. The paper verifies the effectiveness of this method through two examples of fluid - structure interactions, including free - oscillation and forced - oscillation scenarios.
Through this method, the researchers have demonstrated the possibility of reconstructing complex flow fields from limited surface pressure measurement data, which has important application prospects for real - time monitoring and control of fluid - dynamic processes.