Real-time Data-Driven Active Flow Control with Optimal Prediction

This PhD project at The University of Manchester, within the Department of Mechanical and Aerospace Engineering, focuses on developing a real-time, data-driven active flow control framework using optimal prediction methods. Unsteady flows, common in vehicles and buildings, can lead to undesirable aerodynamic loads, noise, and structural fatigue due to low-frequency instabilities interacting with turbulence. Localized flow control techniques, such as jet actuation, are effective in suppressing these instabilities, but the complex, nonlinear nature of turbulent flows makes it challenging to determine optimal control settings in advance. Recent advances in machine learning, especially reinforcement learning, have shown promise in simulation for discovering effective control strategies. However, translating these approaches to real-time experimental settings is difficult due to turbulence, limited sensor coverage, extensive training requirements, and measurement noise. This project aims to overcome these challenges by developing a physics-informed machine learning framework for real-time feedback control of bluff body flows, with a particular focus on suppressing vortex shedding behind a 2D square cylinder. The research will leverage a newly developed data-driven optimal prediction method to inform control decisions from partial measurements, enhancing both efficiency and robustness. Experiments will be conducted in a wind tunnel using a square cylinder equipped with two slot jets for actuation. Aerodynamic drag and surface pressure will be monitored using load cells and pressure sensors. Reduced-order models will be constructed from experimental data to capture the system's response to actuation. Optimal control algorithms will be designed to minimize aerodynamic drag and stabilize wake dynamics in real time, with their effectiveness evaluated during experiments. The project will also assess the scalability and transferability of the control framework to more complex geometries and flow conditions. Eligibility requirements include a minimum of a 2.1 honours degree or a master’s (or international equivalent) in a relevant science or engineering discipline. Applicants must submit transcripts, a CV, a supporting statement outlining motivation and relevant experience, and contact details for two referees. An English language certificate is required if applicable. The application process is online, and early application is recommended as the advert may be removed before the deadline. Funding is available for excellent candidates, covering tuition fees and providing a tax-free stipend at the UKRI rate (£20,780 for 2025/26), with expected annual increases. Self-funded students are also welcome to apply. The start date for the project is October 2026. The university is committed to equality, diversity, and inclusion, and encourages applications from all backgrounds, including those returning from career breaks or seeking flexible study arrangements. For further details and to apply, visit the application portal or contact the admissions team at [email protected]. Prospective applicants are strongly encouraged to contact the supervisor, Prof S Zhong, before applying to discuss their academic background and motivation for the project.