Annika Eichler

Deutsches Elektronen-Synchrotron

Accelerator physics relies on numerical algorithms to solve optimization problems in online accelerator control and tasks such as experimental design and model calibration in simulations. The effectiveness of optimization algorithms in discovering ideal solutions for complex challenges with limited resources often determines the problem complexity...

Bayesian optimization has emerged as a highly effective tool for the safe online optimization of systems, due to its high sample efficiency and noise robustness. To further enhance its efficiency, reduced physical models of the system can be incorporated into the optimization process, accelerating it. These models are able to offer an approximation...

This paper introduces a machine learning-aided fault detection and isolation method applied to the case study of quench identification at the European X-Ray Free-Electron Laser. The plant utilizes 800 superconducting radio-frequency cavities in order to accelerate electron bunches to high energies of up to 17.5 GeV. Various faulty events can disrup...

Online tuning of particle accelerators is a complex optimisation problem that continues to require manual intervention by experienced human operators. Autonomous tuning is a rapidly expanding field of research, where learning-based methods like Bayesian optimisation (BO) hold great promise in improving plant performance and reducing tuning times. A...

Electronic logbooks contain valuable information about activities and events concerning their associated particle accelerator facilities. However, the highly technical nature of logbook entries can hinder their usability and automation. As natural language processing (NLP) continues advancing, it offers opportunities to address various challenges t...

Reinforcement learning (RL) is a unique learning paradigm that is particularly well-suited to tackle complex control tasks, can deal with delayed consequences, and can learn from experience without an explicit model of the dynamics of the problem. These properties make RL methods extremely promising for applications in particle accelerators, where...

Machine learning has emerged as a powerful solution to the modern challenges in accelerator physics. However, the limited availability of beam time, the computational cost of simulations, and the high dimensionality of optimization problems pose significant challenges in generating the required data for training state-of-the-art machine learning mo...

Electronic logbooks contain valuable information about activities and events concerning their associated particle accelerator facilities. However, the highly technical nature of logbook entries can hinder their usability and automation. As natural language processing (NLP) continues advancing, it offers opportunities to address various challenges t...

High-power multi-beam klystrons represent a key component to amplify RF to generate the accelerating field of the superconducting radio frequency (SRF) cavities at European XFEL. Exchanging these high-power components takes time and effort, thus it is necessary to minimize maintenance and downtime and at the same time maximize the device's operatio...

Electronic logbooks contain valuable information about activities and events concerning their associated particle accelerator facilities. However, the highly technical nature of logbook entries can hinder their usability and automation. As natural language processing (NLP) continues advancing, it offers opportunities to address various challenges t...

Autonomous tuning of particle accelerators is an active and challenging field of research with the goal of enabling novel accelerator technologies cutting-edge high-impact applications, such as physics discovery, cancer research and material sciences. A key challenge with autonomous accelerator tuning remains that the most capable algorithms requir...

The successful operation of the laser-based synchronization system of the European X-Ray Free Electron Laser relies on the precise functionality of numerous dynamic systems operating within closed loops with controllers. In this paper, we present how data-based machine learning methods can detect and classify disturbances to such dynamic systems ba...

Machine learning has emerged as a powerful solution to the modern challenges in accelerator physics. However, the limited availability of beam time, the computational cost of simulations, and the high-dimensionality of optimisation problems pose significant challenges in generating the required data for training state-of-the-art machine learning mo...

Online tuning of real-world plants is a complex optimisation problem that continues to require manual intervention by experienced human operators. Autonomous tuning is a rapidly expanding field of research, where learning-based methods, such as Reinforcement Learning-trained Optimisation (RLO) and Bayesian optimisation (BO), hold great promise for...

The optical synchronization system of the European X-ray Free Electron Laser is a networked cyber-physical system producing a large amount of data. To maximize the availability of the optical synchronization system, we are developing a predictive maintenance module that can evaluate and predict the condition of the system. In this paper, we report...

In recent work, it has been shown that reinforcement learning (RL) is capable of outperforming existing methods on accelerator tuning tasks. However, RL algorithms are difficult and time-consuming to train, and currently need to be retrained for every single task. This makes fast deployment in operation difficult and hinders collaborative efforts i...

The problem of robust controller synthesis for plants affected by structured uncertainty, captured by integral quadratic constraints, is discussed. The solution is optimized towards a worst-case white noise rejection specification, which is a generalization of the standard $\mathcal{H}_2$-norm to the robust setting including possibly non-LTI uncert...

A novel approach to detect anomalies in superconducting radio-frequency (rf) cavities is presented, based on the parity space method with the goal to detect quenches and distinguish them from other anomalies. The model-based parity space method relies on analytical redundancy and generates a residual signal computed from measurable rf waveforms. Th...

The European XFEL is currently operating with hundreds of superconducting radio frequency cavities. To be able to minimize the downtimes, prevention of failures on the SRF cavities is crucial. In this paper, we propose an anomaly detection approach based on a neural network model to predict occurrences of breakdowns on the SRF cavities based on a m...

PETRA IV is the upcoming low-emittance, 6 GeV, fourth-generation light source at DESY Hamburg. It is based upon a six-bend achromat lattice with additional beamlines as compared to PETRA III. Stringent stability of the electron beam orbit in the ring will be required to achieve a diffraction-limited photon beam quality. In this regard, the requirem...

A distributed model predictive control scheme is developed for tracking piecewise constant references where the terminal set is reconfigured online, whereas the terminal controller is computed offline. Unlike many standard existing schemes, this scheme yields large feasible regions without performing offline centralized computations. Although the r...

In recent work, it has been shown that reinforcement learning (RL) is capable of solving a variety of problems at sometimes super-human performance levels. But despite continued advances in the field, applying RL to complex real-world control and optimisation problems has proven difficult. In this contribution, we demonstrate how to successfully ap...

Machine learning has proven to be a powerful tool with many applications in the field of accelerator physics. Training machine learning models is a highly iterative process that requires large numbers of samples. However, beam time is often limited and many of the available simulation frameworks are not optimized for fast computation. As a result,...

A novel approach to detect anomalies in superconducting radio-frequency cavities is presented, based on the parity space method with the goal to detect quenches and distinguish them from other anomalies. The model-based parity space method relies on analytical redundancy and generates a residual signal computed from measurable RF waveforms. The res...

Various efforts have been devoted to developing stabilizing distributed Model Predictive Control (MPC) schemes for tracking piecewise constant references. In these schemes, terminal sets are usually computed offline and used in the MPC online phase to guarantee recursive feasibility and asymptotic stability. Maximal invariant terminal sets do not n...

Reinforcement learning algorithms have risen in popularity in the accelerator physics community in recent years, showing potential in beam control and in the optimization and automation of tasks in accelerator operation. The Helmholtz AI project "Machine Learning Toward Autonomous Accelerators" is a collaboration between DESY and KIT that works on...

Superconducting cavities are responsible for beam acceleration in superconducting linear accelerators. Challenging cavity control specifications are necessary to reduce RF costs and to maximize the availability of the accelerator. Cavity detuning and bandwidth are two critical parameters to monitor when operating particle accelerators. Cavity detun...

This paper presents scalable controller synthesis methods for heterogeneous and partially heterogeneous systems. First, heterogeneous systems composed of different subsystems that are interconnected over a directed graph are considered. Techniques from robust and gain-scheduled controller synthesis are employed, in particular the full-block S-proce...

Superconducting accelerating cavities for continuous wave low-current particle accelerators requires a tight resonance control to optimize the RF power costs and to minimize the beam delivery downtime. When the detuning produced by radiation pressure becomes comparable to the RF bandwidth, the monotonic instability starts to affect the cavity opera...

To derive feed-forward signals the impulse response matrix has to be inverted. While for time-invariant systems this matrix has a Toeplitz structure, this is not the case for time-variant systems. Thus, the derivation of the inverse scales cubically with the length of the signal horizon. This paper presents an efficient way to calculate the inverse...

We present an online stochastic model predictive control framework for demand charge management for a grid-connected consumer with attached electrical energy storage. The consumer we consider must satisfy an inflexible but stochastic electricity demand, and also receives a stochastic electricity inflow. The optimization problem formulated solves a...

A novel distributed model predictive control (MPC) scheme is proposed for reference tracking of large-scale systems. In this scheme, the terminal ingredients are reconfigured online taking the current state of the system into account. This results in an infinite-dimensional optimization problem with an infinite number of constraints. By restricting...

In this paper, a novel distributed model predictive control (MPC) scheme with asymmetric adaptive terminal sets is developed for the regulation of large-scale systems with a distributed structure. Similar to typical MPC schemes, a structured Lyapunov matrix and a distributed terminal controller, respecting the distributed structure of the system, a...

This paper presents scalable controller synthesis methods for heterogeneous and partially heterogeneous systems. First, heterogeneous systems composed of different subsystems that are interconnected over a directed graph are considered. Techniques from robust and gain-scheduled controller synthesis are employed, in particular the full-block S-proce...

We present an approximate method for solving nonlinear control problems over long time horizons, in which the full nonlinear model is preserved over an initial part of the horizon, while the remainder of the horizon is modeled using a linear relaxation. As this approximate problem may still be too large to solve directly, we present a Benders decom...

Superconducting accelerating cavities for continuous wave low-current particle accelerators requires a tight resonance control to optimize the RF power costs and to minimize the beam delivery downtime. When the detuning produced by radiation pressure becomes comparable to the RF bandwidth, the monotonic instability starts to affect the cavity opera...

In this paper, a novel distributed model predictive control (MPC) scheme with asymmetric adaptive terminal sets is developed for the regulation of large-scale systems with a distributed structure. Similar to typical MPC schemes, a structured Lyapunov matrix and a distributed terminal controller, respecting the distributed structure of the system, a...

To derive feed-forward signals the impulse response matrix has to be inverted. While for time-invariant systems this matrix has a Toeplitz structure, this is not the case for time-variant systems. Thus, the derivation of the inverse scales cubically with the length of the signal horizon. This paper presents an efficient way to calculate the inverse...

In this paper, we investigate the problem of controlling a seasonal thermal energy storage (STES). The STES considered here is a large scale tank of heated water installed in a building and connected to a solar panel. The stored energy in the STES can be used for providing the building with the space heating (SP) and the domestic hot water (DHW). I...

In this paper, we consider the problem of controller tuning for an operating unit in a building energy system. The illustrative example used here is a real heat pump located in the NEST building at Empa, Dubendorf, Zurich, with its outflow temperature controlled by a PI-controller. The plant is in use and accordingly, intervening in its normal oper...

We present an approximate method for solving nonlinear control problems over long time horizons, in which the full nonlinear model is preserved over an initial part of the horizon, while the remainder of the horizon is modeled using a linear relaxation. As this approximate problem may still be too large to solve directly, we present a Benders decom...

In gene dynamics modeling, parameters of Boolean networks are identified from continuous data under various assumptions expressed by logical constraints. These constraints may restrict the dynamics of the network to the subclass of canalyzing or nested canalyzing functions, which are known to be appropriate for genetic networks. This paper introduc...

In this paper, we propose a distributed model predictive control (DMPC) scheme for linear time-invariant constrained systems which admit a separable structure. To exploit the merits of distributed computation algorithms, the stabilizing terminal controller, value function and invariant terminal set of the DMPC optimization problem need to respect t...

This paper considers the optimization of the convergence speed of consensus under given damping constraints for multi-agent systems with discrete-time double-integrator dynamics with fixed interconnection topology. This work summarizes and details existing results in the case of undirected topologies and extends them to directed ones. The interconn...

Im Kontext der Energiestrategie 2050 des Bundes stellen dezentrale erneuerbare Energiesysteme einen vielver­sprechenden Ansatz dar, da erneuerbare Energie­quellen bereits verfügbar sind. Um technische, ökonomische und soziale Heraus­forderungen solcher Systeme untersuchen zu können, wurden solche Energy-Hubs auf Quartierebene simuliert.

With an increasing share of renewable energy sources, largely decentralized, energy hubs are gaining relevance in the energy landscape as promising solutions because they match local production with consumption. To efficiently control an energy hub a prediction of the energy consumption of the buildings in the hub is required. This work proposes an...

Efficient building energy management has attracted a great deal of academic interest with significant potential energy savings to be envisaged. Social scientists strive to achieve these savings by employing behavior-based approaches, while engineers investigate control strategies for the efficient operation of the building devices. This work can be...

We present a framework for distributed control where the subsystems estimate overlapping components of the state of the overall system. This enables the implementation of decentralized state feedback controllers, which depend on the overlapping state estimates. For a distributed framework, communication can be added. By chosing the amount of commun...

Recent studies in the literature have shown that cooperative energy management of an aggregation of buildings may lead to substantial energy savings. These approaches typically assume the existence of a central operator that is capable of formulating and solving, within a reasonable amount of time, a centralized optimization problem. However, this...

A distributed control scheme is considered, where each subsystem estimates a part of the state space and uses a state feedback-based controller to actuate a subset of the system’s inputs. The estimated parts of the state space are overlapping and thus provide some information about the neighboring subsystems to the local controllers. The number of...

Several studies in the literature have shown the potential energy savings emerging from the cooperative management of the aggregated building energy demands. Sophisticated predictive control schemes have recently been developed that achieve these gains by exploiting the energy generation, conversion and storage equipment shared by the building comm...

This contribution presents and analyzes modeling and minimum cost operation of proton exchange membrane (PEM) fuel cells and electrolyzers. First, detailed thermoelectric models of the electrochemical technologies based on a first-principle approach are presented. Then, as the detailed nonlinear models developed are intractable for use in online op...

This papers considers the design of consensus protocols for discrete-time double-integrator multi-agent systems. It provides solutions for the constrained min-max problem, that optimizes the consensus speed subject to a lower bound on damping. It is proven that the damping is determined by the second smallest and largest eigenvalue of the Laplacian...

In this paper integral quadratic constraints (IQCs) are used for robust stability analysis of interconnected systems with uncertain time-varying interconnection topology, that may encompass uncertain time-varying delays. In contrast to previous results frequency weighted multipliers are used that are known to decrease the conservatism enormously. T...

This paper considers distributed control of a class of interconnected systems, namely decomposable linear parameter-varying (LPV) systems, which include multi-agent systems with LPV agent models and switching communication topology as a special case. Sufficient conditions for stability are established for uncertain time-invariant as well as for tim...

Tensor systems are a framework for modeling of multilinear hybrid systems with discrete and continuous valued signals. Two examples from building services engineering and multi-agent systems show applications of this framework. A tensor model of a heating system is derived and approximated by tensor decomposition methods first. Second, a tensor mod...

This note presents a technique to synthesize distributed controllers for the control of heterogeneous systems interconnected through switching directed interaction topologies. Groups of subsystems are defined with undirected interaction within, but directed interconnections between each other. This allows to construct a virtual symmetric interconne...

This paper considers the convergence speed of multi-agent systems with discrete-time double-integrator dynamics. The communication topology is assumed to be fixed and undirected. The speed of convergence of the associated average consensus protocol is analyzed, and the problem of maximizing the convergence speed over the free parameters in the cons...

This work considers the convergence rate of multi-agent systems with discrete-time single-integrator dynamics and undirected interaction topologies. In recent work it has been proven that in case of lattice interaction topologies the convergence rate can be bounded away from zero, independent of the network size, using asymmetric weightings that gi...

Recently, distributed controller synthesis approaches for decomposable systems, a subclass of distributed systems with identical subsystems, where the interconnection can be described as LFT interconnection, have been proposed. In order to make these approaches tractable for systems containing a very large number of subsystems, constraints on the L...

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 of subsystems are defined with undirected interaction within, but directed interconnections between each other. This allows to construct a...

Recently, distributed robust output feedback control synthesis has been considered for a subclass of interconnected linear parameter-varying (LPV) systems, referred to as decomposable systems. In the synthesis procedure, the interconnection and the LPV uncertainty are described as linear fractional transformations (LFTs), and conditions derived by...

In this paper and its companion paper [1], an approach to distributed control of a class of interconnected systems is proposed. The systems under consideration are known as decomposable systems, which include multi-agent systems as a special case. It is assumed that the interconnection topology is uncertain but belongs to a known set of topologies....

This paper presents a general modeling framework for interconnected LPV systems, that includes model classes like decomposable systems as special cases. The framework allows to consider arbitrary dynamic interconnection operators in the model. We propose to use integral quadratic constraints (IQCs) for robust stability analysis of such interconnect...

In this paper and its companion paper [1], an approach for the distributed control of a class of interconnected systems is proposed. The systems under consideration are known as decomposable systems, which include multi-agent systems as a special case. In this paper, the first part of this series of two papers, sufficient conditions for stability a...

This paper presents convergence bounds for discrete-time second-order multi-agent systems with undirected or directed communication graphs. As has been shown before, the convergence depends on the eigenvalues of the Laplace matrix of the communication graph. For each eigenvalue (or eigenvalue pair) analytic bounds for the parameter set are given to...

This work presents a linear parameter-varying (LPV) approach to distributed control that extends the notion of decomposable systems to decomposable LPV systems. We provide results for the synthesis of distributed output-feedback LPV controllers for heterogeneously scheduled distributed LPV systems in linear fractional (LFT) representation. This res...

In distributed control or for multi-agent systems a decomposition approach often is used for controller synthesis, at which the performance of the decomposed systems is optimized. This work discusses the relationship between the performance of the decomposed and the original system and proposes a method for their optimal calculation. It is shown th...