A5021520570- Advanced Control Systems Optimization
- Stability and Control of Uncertain Systems
- Distributed Control Multi-Agent Systems
- Particle Accelerators and Free-Electron Lasers
- Particle accelerators and beam dynamics
- Fault Detection and Control Systems
- Smart Grid Energy Management
- Control Systems and Identification
- Building Energy and Comfort Optimization
- Neural Networks Stability and Synchronization
- Superconducting Materials and Applications
- Adaptive Control of Nonlinear Systems
- Scientific Computing and Data Management
- Electric Power System Optimization
- Control and Stability of Dynamical Systems
- Medical Imaging Techniques and Applications
- Water resources management and optimization
- Particle Detector Development and Performance
- Magnetic confinement fusion research
- Distributed and Parallel Computing Systems
- Advanced Multi-Objective Optimization Algorithms
- Tensor decomposition and applications
- Oil and Gas Production Techniques
- Mathematical and Theoretical Epidemiology and Ecology Models
- Anomaly Detection Techniques and Applications
Deutsches Elektronen-Synchrotron DESY
2018-2025
Universität Hamburg
2011-2025
Hamburg University of Technology
2011-2025
ETH Zurich
2016-2020
Board of the Swiss Federal Institutes of Technology
2017
Accelerator physics relies on numerical algorithms to solve optimization problems in online accelerator control and tasks such as experimental design model calibration simulations. The effectiveness of discovering ideal solutions for complex challenges with limited resources often determines the problem complexity these methods can address. community has recognized advantages Bayesian algorithms, which leverage statistical surrogate models objective functions effectively address challenges,...
Autonomous tuning of particle accelerators is an active and challenging research field with the goal enabling advanced accelerator technologies cutting-edge high-impact applications, such as physics discovery, cancer research, material sciences. A challenge autonomous remains that most capable algorithms require experts in optimization machine learning to implement them for every new task. Here, we propose use large language models (LLMs) tune accelerators. We demonstrate on a...
Machine learning has emerged as a powerful solution to the modern challenges in accelerator physics. However, limited availability of beam time, computational cost simulations, and high dimensionality optimization problems pose significant generating required data for training state-of-the-art machine models. In this work, we introduce heetah, yorch-based high-speed differentiable linear dynamics code. heetah enables fast collection large datasets by reducing computation times multiple...
Abstract Online tuning of particle accelerators is a complex optimisation problem that continues to require manual intervention by experienced human operators. Autonomous rapidly expanding field research, where learning-based methods like Bayesian (BO) hold great promise in improving plant performance and reducing times. At the same time, reinforcement learning (RL) capable method intelligent controllers, recent work shows RL can also be used train domain-specialised optimisers so-called...
This study investigates the influence of seismic activities on optical synchronization system European X-ray Free-Electron Laser. We analyze controller I/O data phase-locked-loops in length-stabilized links, focusing response to earthquakes, ocean-generated microseism and civilization noise. By comparing with external data, we were able identify disturbances their effects control signals. Our results show that events stability phase-locked loops. Even earthquakes are approximately...
An abstract is not available for this content so a preview has been provided. As you have access to content, full PDF via the 'Save PDF' action button.
Artificial intelligence is increasingly shaping the way we think, design, build and operate present future particle accelerators for science discovery society.
We propose a distributed model predictive control scheme for linear time-invariant constrained systems that admit separable structure. To exploit the merits of computation algorithms, terminal cost and invariant set optimal problem need to respect coupling structure system. Existing methods address this issue typically separate synthesis controllers costs from one sets, do not explicitly consider effect current predicted system states on process. These limitations can adversely affect...
Several studies in the literature have shown potential energy savings emerging from cooperative management of aggregated building demands. Sophisticated predictive control schemes recently been developed that achieve these gains by exploiting generation, conversion, and storage equipment shared community. A common difficulty with all methods is integrating knowledge about long term evolution disturbances affecting system dynamics (e.g., ambient temperature solar radiation). In this context,...
This work presents a linear parameter-varying (LPV) approach to distributed control that extends the notion of decomposable systems LPV systems. We provide results for synthesis output-feedback controllers heterogeneously scheduled in fractional (LFT) representation. result includes multi-agent (MAS) as special case and therefore allows gain-scheduled when each agent's dynamics is dependent on individual parameter values. turn, makes it possbile consider nonlinear agent models if they can be...
This contribution presents and analyzes modeling minimum cost operation of proton exchange membrane (PEM) fuel cells electrolyzers. First, detailed thermoelectric models the electrochemical technologies based on a first-principle approach are presented. Then, as nonlinear developed intractable for use in online optimal control computation, mixed integer linear program (MILP) is formulated with piecewise affine approximation conversion efficiency temperature dynamics devices. The outputs...
In this paper, we consider the problem of controller tuning for an operating unit in a building energy system. As illustrative plant example focus on heat pump. Since is use, method supposed to not intervene with its operation. Moreover, procedure be online, model-free, based only historical data and needs provide safety guarantees regard, formulate as black-box optimization adopt safe Bayesian approaches parameter tuning. These are relatively new control community intensively studied...
Superconducting cavities are responsible for beam acceleration in superconducting linear accelerators. Challenging cavity control specifications necessary to reduce radio frequency (RF) costs and maximize the availability of accelerator. Cavity detuning bandwidth two critical parameters monitor when operating particle is strongly related power required generate desired accelerating gradient. RF losses. A sudden increase can indicate presence a quench or multipacting event. Therefore,...
A technique to synthesize distributed linear parameter-varying (LPV) controllers for the control of heterogeneous LPV systems interconnected through switching directed interaction topologies is presented. Groups subsystems are defined with undirected within, but interconnections between each other. This allows construct a virtual symmetric interconnection matrix representation graph topology. The symmetry guarantees existence diagonalizing transformation, which renders both analysis and...
A novel approach to detect anomalies in superconducting radio-frequency cavities is presented, based on the parity space method with goal quenches and distinguish them from other anomalies. The model-based relies analytical redundancy generates a residual signal computed measurable RF waveforms. sensitive indicator of deviation model provides different signatures for types This new not only helps detecting faults, but also catalogue unique signatures, detected fault. was experimentally...
This article presents scalable controller synthesis methods for heterogeneous and partially systems. First, systems composed of different subsystems that are interconnected over a directed graph considered. Techniques from robust gain-scheduled employed, in particular, the full-block S-procedure, to deal with decentralized system part nominal condition interconnection multiplier condition. Under some structural assumptions, we can decompose conditions into size individual subsystems. To...
Machine learning has emerged as a powerful solution to the modern challenges in accelerator physics. However, limited availability of beam time, computational cost simulations, and high-dimensionality optimisation problems pose significant generating required data for training state-of-the-art machine models. In this work, we introduce Cheetah, PyTorch-based high-speed differentiable linear-beam dynamics code. Cheetah enables fast collection large sets by reducing computation times multiple...
Autonomous tuning of particle accelerators is an active and challenging field research with the goal enabling novel accelerator technologies cutting-edge high-impact applications, such as physics discovery, cancer material sciences. A key challenge autonomous remains that most capable algorithms require expert in optimisation, machine learning or a similar to implement algorithm for every new task. In this work, we propose use large language models (LLMs) tune accelerators. We demonstrate on...
Efficient building energy management has attracted a great deal of academic interest with significant potential savings to be envisaged. Social scientists strive achieve these by employing behavior-based approaches, while engineers investigate control strategies for the efficient operation devices. This work can seen as first step towards bridging two approaches proposing scheme that encapsulates occupant behavior into system. In particular, occupants willingness tolerate comfort bound...
This note presents a technique to synthesize distributed controllers for the control of heterogeneous systems interconnected through switching directed interaction topologies. Groups subsystems are defined with undirected within, but interconnections between each other. allows construct virtual symmetric interconnection matrix representation graph topology. The symmetry guarantees existence diagonalizing transformation, which renders both analysis and synthesis problems particularly simple....
This paper presents a general modeling framework for interconnected LPV systems, that includes model classes like decomposable systems as special cases. The allows to consider arbitrary dynamic interconnection operators in the model. We propose use integral quadratic constraints (IQCs) robust stability analysis of such and provide convex sufficient conditions practically relevant case when communication between subsystems is delayed by an uncertain time-varying time delay. For purpose,...