Pawan Goyal
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Pawan Goyal has a total of 127 co-authors. Here are the top co-authors by number of shared publications:
- Peter Benner
Max Planck Institute for Dynamics of Complex Technical Systems, D-39106 Magdeburg, Germany
46 shared publications - Igor Pontes Duff
Max Planck Institute for Dynamics of Complex Technical Systems Magdeburg Germany
15 shared publications Peter Benner
9 shared publications- Süleyman Yıldız
Max Planck Institute for Dynamics of Complex Technical Systems, 39106, Magdeburg, Germany
6 shared publications Martin Redmann
3 shared publicationsMian Ilyas Ahmad
Max Planck Institute for Dynamics of Complex Technical Systems Sandtorstr. 1 39106 Magdeburg Germany
3 shared publications
Latest Publications
- Open Access
- 16 February 2026
- Published by Taylor & Francis in International Journal of Systems Science
- p. 1-36
Equation discovery: performing sparse regression (SINDy) on the refined analytical gradients
- 16 February 2026
- Published by Taylor & Francis in International Journal of Systems Science
- p. 1-36
Open AccessEquation discovery: performing sparse regression (SINDy) on the refined analytical gradients
- Open Access
- 9 January 2026
- Published by Wiley in International Journal for Numerical Methods in Engineering
- Vol. 127 (1)
Active Sampling of Interpolation Points to Identify Dominant Subspaces for Model Reduction
Model reduction is an active research field to construct low‐dimensional surrogate models of high fidelity to accelerate engineering design cycles. In this work, we investigate model reduction for linear structured systems using dominant reachable and observable subspaces. When the training set—containing all possible interpolation points—is large, these subspaces can be determined by solving many large‐scale linear systems. However, for high‐dimensional models, this easily becomes computationally intractable. To circumvent this issue, in this work, we propose an active sampling strategy to sample only a few points from the given training set, which can allow us to estimate those subspaces accurately. To this end, we formulate the identification of the subspaces as the solution of the generalized Sylvester equations, guiding us to select the most relevant samples from the training set to achieve our goals. Consequently, we construct solutions of the matrix equations in low‐rank forms, which encode subspace information. We extensively discuss computational aspects and efficient usage of the low‐rank factors in the process of obtaining reduced‐order models. We illustrate the proposed active sampling scheme to obtain reduced‐order models via dominant reachable and observable subspaces and present its comparison with the method where all the points from the training set are taken into account. It is shown that the active sample strategy can provide us a speed‐up by one order of magnitude without sacrificing any noticeable accuracy.
- 9 January 2026
- Published by Wiley in International Journal for Numerical Methods in Engineering
- Vol. 127 (1)
Open AccessActive Sampling of Interpolation Points to Identify Dominant Subspaces for Model Reduction
Model reduction is an active research field to construct low‐dimensional surrogate models of high fidelity to accelerate engineering design cycles. In this work, we investigate model reduction for linear structured systems using dominant reachable and observable subspaces. When the training set—containing all possible interpolation points—is large, these subspaces can be determined by solving many large‐scale linear systems. However, for high‐dimensional models, this easily becomes computationally intractable. To circumvent this issue, in this work, we propose an active sampling strategy to sample only a few points from the given training set, which can allow us to estimate those subspaces accurately. To this end, we formulate the identification of the subspaces as the solution of the generalized Sylvester equations, guiding us to select the most relevant samples from the training set to achieve our goals. Consequently, we construct solutions of the matrix equations in low‐rank forms, which encode subspace information. We extensively discuss computational aspects and efficient usage of the low‐rank factors in the process of obtaining reduced‐order models. We illustrate the proposed active sampling scheme to obtain reduced‐order models via dominant reachable and observable subspaces and present its comparison with the method where all the points from the training set are taken into account. It is shown that the active sample strategy can provide us a speed‐up by one order of magnitude without sacrificing any noticeable accuracy.
- Top Cited PapersOpen Access
- 10 September 2025
- Published by Elsevier in Physica D: Nonlinear Phenomena
- Vol. 483
Guaranteed stable quadratic models and their applications in SINDy and operator inference
- 10 September 2025
- Published by Elsevier in Physica D: Nonlinear Phenomena
- Vol. 483
Top Cited PapersOpen AccessGuaranteed stable quadratic models and their applications in SINDy and operator inference
- 1 August 2025
- Published by Walter de Gruyter GmbH in at - Automatisierungstechnik
- Vol. 73 (8) , 572-591
Approximate balanced truncation for linear structured systems based on greedy numerical integration
This paper studies balanced truncation for linear structured system (LSS). We focus on the efficient computation of controllability and observability Gramians. Unlike standard linear time-invariant (LTI) systems, there is no general algebraic Lyapunov equation that encodes the Gramians for LSS. Instead, natural approaches approximate the Gramians using quadrature rules, often requiring many nodes. We identify these approximations with the solutions to generalized Sylvester equations, allowing us to employ and adapt an existing active sampling strategy and compute low-rank factors of the Gramians as solutions to the matrix equations. Numerical examples show comparable accuracy to standard methods with similar computational time for medium-sized systems, and speed-up and reduced memory requirements for large dynamical systems.
- 1 August 2025
- Published by Walter de Gruyter GmbH in at - Automatisierungstechnik
- Vol. 73 (8) , 572-591
Approximate balanced truncation for linear structured systems based on greedy numerical integration
This paper studies balanced truncation for linear structured system (LSS). We focus on the efficient computation of controllability and observability Gramians. Unlike standard linear time-invariant (LTI) systems, there is no general algebraic Lyapunov equation that encodes the Gramians for LSS. Instead, natural approaches approximate the Gramians using quadrature rules, often requiring many nodes. We identify these approximations with the solutions to generalized Sylvester equations, allowing us to employ and adapt an existing active sampling strategy and compute low-rank factors of the Gramians as solutions to the matrix equations. Numerical examples show comparable accuracy to standard methods with similar computational time for medium-sized systems, and speed-up and reduced memory requirements for large dynamical systems.
- 25 April 2025
- Published by Springer Nature
- p. 365-376
Learning Reduced-Order Quadratic-Linear Models in Process Engineering Using Operator Inference
- 25 April 2025
- Published by Springer Nature
- p. 365-376
Learning Reduced-Order Quadratic-Linear Models in Process Engineering Using Operator Inference
- Open Access
- 9 April 2025
- Published by Springer Nature in Advanced Modeling and Simulation in Engineering Sciences
- Vol. 12 (1) , 1-19
Structure-preserving learning for multi-symplectic PDEs
This paper presents an energy-preserving machine learning method for inferring reduced-order models (ROMs) by exploiting the multi-symplectic form of partial differential equations (PDEs). The vast majority of energy-preserving reduced-order methods use symplectic Galerkin projection to construct reduced-order Hamiltonian models by projecting the full models onto a symplectic subspace. However, symplectic projection requires the existence of fully discrete operators, and in many cases, such as black-box PDE solvers, these operators are inaccessible. In this work, we propose an energy-preserving machine learning method that can infer the dynamics of the given PDE using data only, so that the proposed framework does not depend on the fully discrete operators. In this context, the proposed method is non-intrusive. The proposed method is grey box in the sense that it requires only some basic knowledge of the multi-symplectic model at the partial differential equation level. We prove that the proposed method satisfies spatially discrete local energy conservation and preserves the multi-symplectic conservation laws. We test our method on the linear wave equation, the Korteweg-de Vries equation, and the Zakharov-Kuznetsov equation. We test the generalization of our learned models by testing them far outside the training time interval.
- 9 April 2025
- Published by Springer Nature in Advanced Modeling and Simulation in Engineering Sciences
- Vol. 12 (1) , 1-19
Open AccessStructure-preserving learning for multi-symplectic PDEs
This paper presents an energy-preserving machine learning method for inferring reduced-order models (ROMs) by exploiting the multi-symplectic form of partial differential equations (PDEs). The vast majority of energy-preserving reduced-order methods use symplectic Galerkin projection to construct reduced-order Hamiltonian models by projecting the full models onto a symplectic subspace. However, symplectic projection requires the existence of fully discrete operators, and in many cases, such as black-box PDE solvers, these operators are inaccessible. In this work, we propose an energy-preserving machine learning method that can infer the dynamics of the given PDE using data only, so that the proposed framework does not depend on the fully discrete operators. In this context, the proposed method is non-intrusive. The proposed method is grey box in the sense that it requires only some basic knowledge of the multi-symplectic model at the partial differential equation level. We prove that the proposed method satisfies spatially discrete local energy conservation and preserves the multi-symplectic conservation laws. We test our method on the linear wave equation, the Korteweg-de Vries equation, and the Zakharov-Kuznetsov equation. We test the generalization of our learned models by testing them far outside the training time interval.
- Open Access
- 13 March 2025
- Published by IOP Publishing in Machine Learning: Science and Technology
- Vol. 6 (1) , 015063
Deep learning for structure-preserving universal stable Koopman-inspired embeddings for nonlinear canonical Hamiltonian dynamics
Discovering a suitable coordinate transformation for nonlinear systems enables the construction of simpler models, facilitating prediction, control, and optimization for complex nonlinear systems. To that end, Koopman operator theory offers a framework for global linearization of nonlinear systems, thereby allowing the usage of linear tools for design studies. In this work, we focus on the identification of global linearized embeddings for canonical nonlinear Hamiltonian systems through a symplectic transformation. While this task is often challenging, we leverage the power of deep learning to discover the desired embeddings. Furthermore, to overcome the shortcomings of Koopman operators for systems with continuous spectra, we apply the lifting principle and learn global cubicized embeddings. Additionally, a key emphasis is given to enforce the bounded stability for the dynamics of the discovered embeddings. We demonstrate the capabilities of deep learning in acquiring compact symplectic coordinate transformations and the corresponding simple dynamical models, fostering data-driven learning of nonlinear canonical Hamiltonian systems, even those with continuous spectra.
- 13 March 2025
- Published by IOP Publishing in Machine Learning: Science and Technology
- Vol. 6 (1) , 015063
Open AccessDeep learning for structure-preserving universal stable Koopman-inspired embeddings for nonlinear canonical Hamiltonian dynamics
Discovering a suitable coordinate transformation for nonlinear systems enables the construction of simpler models, facilitating prediction, control, and optimization for complex nonlinear systems. To that end, Koopman operator theory offers a framework for global linearization of nonlinear systems, thereby allowing the usage of linear tools for design studies. In this work, we focus on the identification of global linearized embeddings for canonical nonlinear Hamiltonian systems through a symplectic transformation. While this task is often challenging, we leverage the power of deep learning to discover the desired embeddings. Furthermore, to overcome the shortcomings of Koopman operators for systems with continuous spectra, we apply the lifting principle and learn global cubicized embeddings. Additionally, a key emphasis is given to enforce the bounded stability for the dynamics of the discovered embeddings. We demonstrate the capabilities of deep learning in acquiring compact symplectic coordinate transformations and the corresponding simple dynamical models, fostering data-driven learning of nonlinear canonical Hamiltonian systems, even those with continuous spectra.
- Open Access
Structure-preserving learning for multi-symplectic PDEs
This paper presents an energy-preserving machine learning method for inferring reduced-order models (ROMs) by exploiting the multi-symplectic form of partial differential equations (PDEs). The vast majority of energy-preserving reduced-order methods use symplectic Galerkin projection to construct reduced-order Hamiltonian models by projecting the full models onto a symplectic subspace. However, symplectic projection requires the existence of fully discrete operators, and in many cases, such as black-box PDE solvers, these operators are inaccessible. In this work, we propose an energy-preserving machine learning method that can infer the dynamics of the given PDE using data only, so that the proposed framework does not depend on the fully discrete operators. In this context, the proposed method is non-intrusive. The proposed method is grey box in the sense that it requires only some basic knowledge of the multi-symplectic model at the partial differential equation level. We prove that the proposed method satisfies spatially discrete local energy conservation and preserves the multi-symplectic conservation laws. We test our method on the linear wave equation, the Korteweg-de Vries equation, and the Zakharov-Kuznetsov equation. We test the generalization of our learned models by testing them far outside the training time interval.
Open AccessStructure-preserving learning for multi-symplectic PDEs
This paper presents an energy-preserving machine learning method for inferring reduced-order models (ROMs) by exploiting the multi-symplectic form of partial differential equations (PDEs). The vast majority of energy-preserving reduced-order methods use symplectic Galerkin projection to construct reduced-order Hamiltonian models by projecting the full models onto a symplectic subspace. However, symplectic projection requires the existence of fully discrete operators, and in many cases, such as black-box PDE solvers, these operators are inaccessible. In this work, we propose an energy-preserving machine learning method that can infer the dynamics of the given PDE using data only, so that the proposed framework does not depend on the fully discrete operators. In this context, the proposed method is non-intrusive. The proposed method is grey box in the sense that it requires only some basic knowledge of the multi-symplectic model at the partial differential equation level. We prove that the proposed method satisfies spatially discrete local energy conservation and preserves the multi-symplectic conservation laws. We test our method on the linear wave equation, the Korteweg-de Vries equation, and the Zakharov-Kuznetsov equation. We test the generalization of our learned models by testing them far outside the training time interval.
- Open Access
Active Sampling of Interpolation Points to Identify Dominant Subspaces for Model Reduction
Model reduction is an active research field to construct low-dimensional surrogate models of high fidelity to accelerate engineering design cycles. In this work, we investigate model reduction for linear structured systems using dominant reachable and observable subspaces. When the training set $-$ containing all possible interpolation points $-$ is large, then these subspaces can be determined by solving many large-scale linear systems. However, for high-fidelity models, this easily becomes computationally intractable. To circumvent this issue, in this work, we propose an active sampling strategy to sample only a few points from the given training set, which can allow us to estimate those subspaces accurately. To this end, we formulate the identification of the subspaces as the solution of the generalized Sylvester equations, guiding us to select the most relevant samples from the training set to achieve our goals. Consequently, we construct solutions of the matrix equations in low-rank forms, which encode subspace information. We extensively discuss computational aspects and efficient usage of the low-rank factors in the process of obtaining reduced-order models. We illustrate the proposed active sampling scheme to obtain reduced-order models via dominant reachable and observable subspaces and present its comparison with the method where all the points from the training set are taken into account. It is shown that the active sample strategy can provide us $17$x speed-up without sacrificing any noticeable accuracy.Open AccessActive Sampling of Interpolation Points to Identify Dominant Subspaces for Model Reduction
Model reduction is an active research field to construct low-dimensional surrogate models of high fidelity to accelerate engineering design cycles. In this work, we investigate model reduction for linear structured systems using dominant reachable and observable subspaces. When the training set $-$ containing all possible interpolation points $-$ is large, then these subspaces can be determined by solving many large-scale linear systems. However, for high-fidelity models, this easily becomes computationally intractable. To circumvent this issue, in this work, we propose an active sampling strategy to sample only a few points from the given training set, which can allow us to estimate those subspaces accurately. To this end, we formulate the identification of the subspaces as the solution of the generalized Sylvester equations, guiding us to select the most relevant samples from the training set to achieve our goals. Consequently, we construct solutions of the matrix equations in low-rank forms, which encode subspace information. We extensively discuss computational aspects and efficient usage of the low-rank factors in the process of obtaining reduced-order models. We illustrate the proposed active sampling scheme to obtain reduced-order models via dominant reachable and observable subspaces and present its comparison with the method where all the points from the training set are taken into account. It is shown that the active sample strategy can provide us $17$x speed-up without sacrificing any noticeable accuracy. - Open Access
- 8 August 2024
- Published by Springer Nature in Advances in Computational Mathematics
- Vol. 50 (4) , 1-31
Balanced truncation for quadratic-bilinear control systems
We discuss model order reduction (MOR) for large-scale quadratic-bilinear (QB) systems based on balanced truncation. The method for linear systems mainly involves the computation of the Gramians of the system, namely reachability and observability Gramians. These Gramians are extended to a general nonlinear setting in Scherpen (Systems Control Lett. 21, 143-153 1993). These formulations of Gramians are not only challenging to compute for large-scale systems but hard to utilize also in the MOR framework. This work proposes algebraic Gramians for QB systems based on the underlying Volterra series representation of QB systems and their Hilbert adjoint systems. We then show their relation to a certain type of generalized quadratic Lyapunov equation. Furthermore, we quantify the reachability and observability subspaces based on the proposed Gramians. Consequently, we propose a balancing algorithm, allowing us to find those states that are simultaneously hard to reach and hard to observe. Truncating such states yields reduced-order systems. We also study sufficient conditions for the existence of Gramians, and a local stability of reduced-order models obtained using the proposed balanced truncation scheme. Finally, we demonstrate the proposed balancing-type MOR for QB systems using various numerical examples.
- 8 August 2024
- Published by Springer Nature in Advances in Computational Mathematics
- Vol. 50 (4) , 1-31
Open AccessBalanced truncation for quadratic-bilinear control systems
We discuss model order reduction (MOR) for large-scale quadratic-bilinear (QB) systems based on balanced truncation. The method for linear systems mainly involves the computation of the Gramians of the system, namely reachability and observability Gramians. These Gramians are extended to a general nonlinear setting in Scherpen (Systems Control Lett. 21, 143-153 1993). These formulations of Gramians are not only challenging to compute for large-scale systems but hard to utilize also in the MOR framework. This work proposes algebraic Gramians for QB systems based on the underlying Volterra series representation of QB systems and their Hilbert adjoint systems. We then show their relation to a certain type of generalized quadratic Lyapunov equation. Furthermore, we quantify the reachability and observability subspaces based on the proposed Gramians. Consequently, we propose a balancing algorithm, allowing us to find those states that are simultaneously hard to reach and hard to observe. Truncating such states yields reduced-order systems. We also study sufficient conditions for the existence of Gramians, and a local stability of reduced-order models obtained using the proposed balanced truncation scheme. Finally, we demonstrate the proposed balancing-type MOR for QB systems using various numerical examples.