Skip to content
Preprint

Identification of effective transport physics and optimal feedback control: Experimental demonstration on a 1600-fan wind generator

Oct 2026 · 0 citations · 32 references
Physics

Abstract

Distributed-input distributed-output (DIDO) systems arise in numerous engineering applications where large numbers of actuators generate spatially distributed physical fields that are monitored by distributed sensors. Developing physically interpretable models that enable real-time optimal control of these systems remains challenging because of the high dimensionality and strong spatial coupling between inputs and outputs. In this work, we propose a physics-based framework for modeling and closed-loop optimal control of DIDO systems from experimental input-output data. First, an effective partial differential equation (PDE) is selected among candidate transport models and discretized on the actuator grid to obtain a minimum-grid representation. Second, the PDE is solved in closed form, which yields an input-output operator with only four physically interpretable parameters. Third, the operator is incorporated into an iterative, constrained quadratic program that accounts for the actuator limits. The framework is demonstrated on the inner 40 x 40 section of the world's highest-resolution Fan Array Wind Generator (FAWG), comprising 1600 independently controlled fans with 36 distributed velocity measurements. The model describes the spatial interaction between the 1600 actuators and the measured velocity field with normalized mean absolute prediction errors of 11-20% for representative actuation patterns. The closed-loop controller reproduces all three prescribed two-dimensional velocity fields with a normalized mean absolute error below 2.3% in fewer than 10 iterations.

View source

Similar papers

Preprint Sep 2026

Port-Hamiltonian Koopman Operator Synthesis for Mechanical Systems

This work presents a structure-preserving Koopman framework for Euler-Lagrange systems built on generalized-momentum coordinates, and demonstrates improved prediction accuracy, data efficiency, and closed-loop tracking over Koopman baselines, with increasing gains for higher-dimensional systems.

Rajpal Singh, Aditya Singh, J. Keshavan · 0 citations
Open access Sep 2026

Artificial Intelligence Reduced Order Observer Design for Output Feedback Control

In modern control engineering, the rising demand for streamlined state estimation in dynamic systems has made the development of reduced-order observers a critical area of study. This paper explores the integration of neural networks into observer design, utilizing their specialized ability to learn unknown dynamics an...

O. Alsmadi, Zaer Abu-Hammour, Feras Alasali · 0 citations
Review Open access Oct 2026

Feedback control systems for plasma operation at Wendelstein 7-X

Real-time feedback control is essential for meeting the physics and operational demands of Wendelstein 7-X (W7-X), particularly the need to respond effectively to transient events, maintain stationary high-performance phases and operate reliably across a wide range of magnetic configurations and plasma regimes. This wo...

Thomas Wegner, M. Krychowiak, L. Krier · 0 citations
#machine learning Preprint Nov 2026

Rapid Mean-Field Formation Transport with a Shared-Representation Neural Fokker-Planck Controller

This paper develops a mean-field formation transport method with low online computational cost for large-scale multi-agent systems. Classical mean-field planners often require repeated solutions of coupled forward and backward partial differential equations, making online replanning computationally demanding. Density a...

Ze-Jian Zhou, Hao Xu · 0 citations
Open access Nov 2026

A two-step physics-informed neural network approach for open-loop optimal control of the Gardner equation with limited data and partially unknown parameters

This paper introduces a two-stage PINN framework, using a Lagrangian multiplier framework, and employs PINNs to jointly approximate the state, adjoint, and control variables, enabling effective learning even under severe data scarcity.

Sana Idrees, Shahbaz Ahmad · 0 citations
Preprint Oct 2026

Data-Driven Static Output-Feedback Control for Multi-Agent Systems

This work proposes a data-driven framework that synthesizes output-feedback control for linear time-invariant multi-agent systems (MAS). We first develop two algorithms that use offline data samples of state and input trajectories to determine a Nash equilibrium (NE) for MAS. Next, we formulate a semidefinite program t...

Paul Serna-Torre, Adam B. Sedlak, Patricia Hidalgo-Gonzalez · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.