Load estimation in unsteady flows from sparse pressure measurements: Application of transition networks to experimental data

Abstract

Inspired by biological swimming and flying with distributed sensing, we propose a data-driven approach for load estimation that relies on complex networks. We exploit sparse, real-time pressure inputs, combined with pre-trained transition networks, to estimate aerodynamic loads in unsteady and highly separated flows. The transition networks contain the aerodynamic states of the system as nodes along with the underlying dynamics as links. A weighted average-based (WAB) strategy is proposed and tested on realistic experimental data on the flow around an accelerating elliptical plate at various angles of attack. Aerodynamic loads are then estimated for angles-of-attack cases not included in the training dataset so as to simulate the estimation process. An optimization process is also included to account for the system's temporal dynamics. Performance and limitations of the WAB approach are discussed, showing that transition networks can represent a versatile and effective data-driven tool for real-time signal estimation using sparse and noisy signals (such as surface pressure) in realistic flows.

Cluster-based Bayesian approach for noisy and sparse data: application to flow-state estimation

“…For example, as shown in figure 4c, states with high noise levels can be assigned large probabilities ( E i ) during signal estimation despite being significantly different from the measured input m(t). This feature distinguishes the CBM approach from previous work [14] as visualized in figure 4d,e. Moreover, for the estimation stage, time series for a randomly selected individual run were used to simulate a realistic scenario since, in practice, noisy individual signals (i.e.…”

“…An exemplifying sketch illustrating these three challenges for an unsteady, bioinspired, aerodynamic problem is shown in figure 1. While previous attempts to address sensor sparsity for flow estimation have been conducted [12,14,17], to date the role of measurement uncertainty has often only been investigated by adding artificial noise (e.g. uniformly sampled noise) to noise-free data [19].…”

“…In this study, we propose a data-driven approach for real-time flow (aerodynamic) state estimation that is conceived for realistic scenarios involving noisy, sparse and limited data. Specifically, we extend the previous cluster-based methodologies relying on transition networks [14,20] through a Bayesian estimation process that explicitly accounts for the presence of Ten runs are shown to demonstrate some challenges of experimental data: sparsity, limited data, dynamic response of the measurement system, run-to-run scatter and state-dependent noise levels; (d) a data-driven model able to estimate the system's state and unmeasured variables using real-time input of a subset of sensors.…”

Bayesian inference of vorticity in unbounded flow from limited pressure measurements

“…We focus in this paper on the inference of incompressible flows of moderate and large Reynolds numbers from pressure sensors, a problem that has been of interest in the fluid dynamics community for many years (Naguib, Wark & Juckenhöfel 2001; Murray & Ukeiley 2003; Gomez et al. 2019; Sashittal & Bodony 2021; Iacobello, Kaiser & Rival 2022; Zhong et al. 2023).…”

“…The problem of flow estimation is very broad and can be pursued with different types of sensors in the presence of various flow physics. We focus in this paper on the inference of incompressible flows of moderate and large Reynolds numbers from pressure sensors, a problem that has been of interest in the fluid dynamics community for many years (Naguib, Wark & Juckenhöfel 2001;Murray & Ukeiley 2003;Gomez et al 2019;Sashittal & Bodony 2021;Iacobello, Kaiser & Rival 2022;Zhong et al 2023). Flow estimation from other types of noisy measurements has also been pursued in closely related contexts in recent years, with tools very similar to those used in the present work (Juniper & Yoko 2022;Kontogiannis et al 2022).…”