Pavel V. Kuptsov
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Pavel V. Kuptsov has a total of 32 co-authors. Here are the top co-authors by number of shared publications:
Sergey P. Kuznetsov
Kotel'nikov's Institute of Radio-Engineering and Electronics of RAS Saratov Branch Zelenaya 38 410019 Saratov Russia
13 shared publicationsAnna V. Kuptsova
Institute of electronics and mechanical engineering Yuri Gagarin State Technical University of Saratov Politekhnicheskaya 77 410054 Saratov Russia
12 shared publications- Nataliya V. Stankevich
HSE University, Nizhny Novgorod, Russia
8 shared publications Ulrich Parlitz
Biomedical Physics Group Max Planck Institute for Dynamics and Self-Organization Am Faßberg 17 37077 Göttingen Germany
4 shared publicationsRazvan A. Satnoianu
3 shared publicationsArkady Pikovsky
Institute of Physics and Astronomy University of Potsdam Karl-Liebknecht-Str. 24/25 14476 Potsdam-Golm Germany
2 shared publications
Latest Publications
- 1 May 2026
- Published by AIP Publishing in Chaos: An Interdisciplinary Journal of Nonlinear Science
- Vol. 36 (5)
Neural-network maps for two-parameter modeling of bistability and codimension-two bifurcations in two-dimensional flow dynamical systems
We develop a machine-learning approach to reproduce the behavior of two versions of the van der Pol oscillator exhibiting a subcritical Andronov–Hopf bifurcation, with or without a codimension-2 Bautin point. We construct a neural-network model that functions as a recurrent map and train it on short segments of oscillator trajectories. The results show that, despite its relatively simple architecture of neural network, the trained map is capable of capturing the key dynamical features of the original oscillators, even though corresponding examples are absent from the training data. In a two-dimensional parameter space, the neural-network map reproduces, without direct exposure during training, the multistability of the oscillators and their bifurcations, including the Andronov–Hopf bifurcation, the saddle-node bifurcation associated with the birth of a limit cycle, and the codimension-2 Bautin bifurcation. We further show that, although explicit examples of these effects are not necessary, the ability to recover them still depends on an appropriate choice of training data.
- 1 May 2026
- Published by AIP Publishing in Chaos: An Interdisciplinary Journal of Nonlinear Science
- Vol. 36 (5)
Neural-network maps for two-parameter modeling of bistability and codimension-two bifurcations in two-dimensional flow dynamical systems
We develop a machine-learning approach to reproduce the behavior of two versions of the van der Pol oscillator exhibiting a subcritical Andronov–Hopf bifurcation, with or without a codimension-2 Bautin point. We construct a neural-network model that functions as a recurrent map and train it on short segments of oscillator trajectories. The results show that, despite its relatively simple architecture of neural network, the trained map is capable of capturing the key dynamical features of the original oscillators, even though corresponding examples are absent from the training data. In a two-dimensional parameter space, the neural-network map reproduces, without direct exposure during training, the multistability of the oscillators and their bifurcations, including the Andronov–Hopf bifurcation, the saddle-node bifurcation associated with the birth of a limit cycle, and the codimension-2 Bautin bifurcation. We further show that, although explicit examples of these effects are not necessary, the ability to recover them still depends on an appropriate choice of training data.
- Open Access
An improved approach for estimating the dimension of inertial manifolds in chaotic distributed dynamical systems via analysis of angles between tangent subspaces
While a previously proposed method for estimating inertial manifold dimension, based on explicitly computing angles between pairs of covariant Lyapunov vectors (CLVs), employs efficient algorithms, it remains computationally demanding due to its substantial resource requirements. In this work, we introduce an improved method to determine this dimension by analyzing the angles between tangent subspaces spanned by the CLVs. This approach builds upon a fast numerical technique for assessing chaotic dynamics hyperbolicity. Crucially, the proposed method requires significantly less computational effort and minimizes memory usage by eliminating the need for explicit CLV computation. We test our method on two canonical systems: the complex Ginzburg-Landau equation and a diffusively coupled chain of Lorenz oscillators. For the former, the results confirm the accuracy of the new approach by matching prior dimension estimates. For the latter, the analysis demonstrates the absence of a low-dimensional inertial manifold, highlighting a complex regime that merits further investigation. The presented method offers a practical and efficient tool for characterizing high-dimensional attractors in extended dynamical systems.
Open AccessAn improved approach for estimating the dimension of inertial manifolds in chaotic distributed dynamical systems via analysis of angles between tangent subspaces
While a previously proposed method for estimating inertial manifold dimension, based on explicitly computing angles between pairs of covariant Lyapunov vectors (CLVs), employs efficient algorithms, it remains computationally demanding due to its substantial resource requirements. In this work, we introduce an improved method to determine this dimension by analyzing the angles between tangent subspaces spanned by the CLVs. This approach builds upon a fast numerical technique for assessing chaotic dynamics hyperbolicity. Crucially, the proposed method requires significantly less computational effort and minimizes memory usage by eliminating the need for explicit CLV computation. We test our method on two canonical systems: the complex Ginzburg-Landau equation and a diffusively coupled chain of Lorenz oscillators. For the former, the results confirm the accuracy of the new approach by matching prior dimension estimates. For the latter, the analysis demonstrates the absence of a low-dimensional inertial manifold, highlighting a complex regime that merits further investigation. The presented method offers a practical and efficient tool for characterizing high-dimensional attractors in extended dynamical systems.
- 7 April 2025
- Published by Pleiades Publishing Ltd in Regular and Chaotic Dynamics
- Vol. 30 (2) , 291-305
Verification of Chaos in a Human Cardiovascular System Model
This study discusses an approach for estimation of the largest Lyapunov exponent for the mathematical model of the cardiovascular system. The accuracy was verified using the confidence intervals approach. The algorithm was used to investigate the effects of noises with different amplitudes and spectral compositions on the dynamics of the model. Three sets of parameters are considered, corresponding to different states of the human cardiovascular system model. It is shown that, in each case, the model exhibited chaotic dynamics. The model gave different responses to the changes in the characteristics of the noise, when using different sets of parameters. The noise had both constructive and destructive effects, depending on the parameters of the model and the noise, by, respectively, amplifying or inhibiting the chaotic dynamics of the model.
- 7 April 2025
- Published by Pleiades Publishing Ltd in Regular and Chaotic Dynamics
- Vol. 30 (2) , 291-305
Verification of Chaos in a Human Cardiovascular System Model
This study discusses an approach for estimation of the largest Lyapunov exponent for the mathematical model of the cardiovascular system. The accuracy was verified using the confidence intervals approach. The algorithm was used to investigate the effects of noises with different amplitudes and spectral compositions on the dynamics of the model. Three sets of parameters are considered, corresponding to different states of the human cardiovascular system model. It is shown that, in each case, the model exhibited chaotic dynamics. The model gave different responses to the changes in the characteristics of the noise, when using different sets of parameters. The noise had both constructive and destructive effects, depending on the parameters of the model and the noise, by, respectively, amplifying or inhibiting the chaotic dynamics of the model.
- Open Access
- 14 February 2025
- Published by Bryansk State Technical University BSTU in Science intensive technologies in mechanical engineering
- Vol. 2025 (4) , 24-31
Technology and results of studying the diversification of the operative conditions of metal cutting tools
The technology and the results of its investigation contributing to studying the diversification of the operative conditions of metal cutting tools, are presented as exemplified in a tool equipped with a hard alloy plate. It is shown that the rate of diversification can be judged by the appearance of a two-dimensional histogram constructed from time series of signals indicating fluctuations in the technological system in mutually perpendicular planes. The occurrence of a "comet tail" pattern on the histogram indicates a high rate of diversification of operative conditions, in particular, during the running-in of the tool. To show the dynamics of the diversification, a cross-correlation curve is used, the form of which characterizes the strength of the relationship between the directions of oscillation. This networking proves regulation of operative conditions and the transition to a starting operation in a stationary mode. Diversification dynamics is identified using a procedure based on statistical test of the hypothesis for the significance of changes in the transformation coefficients of the spectra of diverse oscillation signals using the Student's t-test. The identification results showed that state diversification dynamics is a change in connections and constraint in the industrial process system, which leads to the formation of new dynamic properties that reduce cutting forces due to the redistribution of friction forces between the front and rear surfaces of the tool and, as a result, increase stability for disturbance resulted from the operating procedure. The approach used in the development of the process desigh is compared with approaches based on the construction of a mathematical (analytical or stochastic) model of a real system. The direction of practical implementation of the technology in the software of CNC systems is justified for more efficient solution of traditional control tasks, in particular, process tasks, which will ensure an increase in the reliability of the tool.
- 14 February 2025
- Published by Bryansk State Technical University BSTU in Science intensive technologies in mechanical engineering
- Vol. 2025 (4) , 24-31
Open AccessTechnology and results of studying the diversification of the operative conditions of metal cutting tools
The technology and the results of its investigation contributing to studying the diversification of the operative conditions of metal cutting tools, are presented as exemplified in a tool equipped with a hard alloy plate. It is shown that the rate of diversification can be judged by the appearance of a two-dimensional histogram constructed from time series of signals indicating fluctuations in the technological system in mutually perpendicular planes. The occurrence of a "comet tail" pattern on the histogram indicates a high rate of diversification of operative conditions, in particular, during the running-in of the tool. To show the dynamics of the diversification, a cross-correlation curve is used, the form of which characterizes the strength of the relationship between the directions of oscillation. This networking proves regulation of operative conditions and the transition to a starting operation in a stationary mode. Diversification dynamics is identified using a procedure based on statistical test of the hypothesis for the significance of changes in the transformation coefficients of the spectra of diverse oscillation signals using the Student's t-test. The identification results showed that state diversification dynamics is a change in connections and constraint in the industrial process system, which leads to the formation of new dynamic properties that reduce cutting forces due to the redistribution of friction forces between the front and rear surfaces of the tool and, as a result, increase stability for disturbance resulted from the operating procedure. The approach used in the development of the process desigh is compared with approaches based on the construction of a mathematical (analytical or stochastic) model of a real system. The direction of practical implementation of the technology in the software of CNC systems is justified for more efficient solution of traditional control tasks, in particular, process tasks, which will ensure an increase in the reliability of the tool.
- 30 November 2024
- Published by Elsevier in Chaos, Solitons & Fractals
- Vol. 191
Reconstruction of neuromorphic dynamics from a single scalar time series using variational autoencoder and neural network map
- 30 November 2024
- Published by Elsevier in Chaos, Solitons & Fractals
- Vol. 191
Reconstruction of neuromorphic dynamics from a single scalar time series using variational autoencoder and neural network map
- Open Access
Reconstruction of neuromorphic dynamics from a single scalar time series using variational autoencoder and neural network map
This paper examines the reconstruction of a family of dynamical systems with neuromorphic behavior using a single scalar time series. A model of a physiological neuron based on the Hodgkin-Huxley formalism is considered. Single time series of one of its variables is shown to be enough to train a neural network that can operate as a discrete time dynamical system with one control parameter. The neural network system is created in two steps. First, the delay-coordinate embedding vectors are constructed form the original time series and their dimension is reduced with by means of a variational autoencoder to obtain the recovered state-space vectors. It is shown that an appropriate reduced dimension can be determined by analyzing the autoencoder training process. Second, pairs of the recovered state-space vectors at consecutive time steps supplied with a constant value playing the role of a control parameter are used to train another neural network to make it operate as a recurrent map. The regimes of thus created neural network system observed when its control parameter is varied are in very good accordance with those of the original system, though they were not explicitly presented during training.
Open AccessReconstruction of neuromorphic dynamics from a single scalar time series using variational autoencoder and neural network map
This paper examines the reconstruction of a family of dynamical systems with neuromorphic behavior using a single scalar time series. A model of a physiological neuron based on the Hodgkin-Huxley formalism is considered. Single time series of one of its variables is shown to be enough to train a neural network that can operate as a discrete time dynamical system with one control parameter. The neural network system is created in two steps. First, the delay-coordinate embedding vectors are constructed form the original time series and their dimension is reduced with by means of a variational autoencoder to obtain the recovered state-space vectors. It is shown that an appropriate reduced dimension can be determined by analyzing the autoencoder training process. Second, pairs of the recovered state-space vectors at consecutive time steps supplied with a constant value playing the role of a control parameter are used to train another neural network to make it operate as a recurrent map. The regimes of thus created neural network system observed when its control parameter is varied are in very good accordance with those of the original system, though they were not explicitly presented during training.
- Open Access
- 10 December 2023
- Published by Saratov State University in Izvestiya VUZ. Applied Nonlinear Dynamics
- Vol. 32 (1) , 72-95
Modeling of the Hodgkin–Huxley neural oscillators dynamics using an artificial neural network
The purpose of this study — to represent a detailed description of the procedure for creating and training a neural network mapping on the example of the dynamics modeling of a neural oscillator of the Hodgkin–Huxley type; to show that the neural network mappings trained for single oscillators can be used as elements of a coupled system that simulate the behavior of coupled oscillators. Methods. Numerical method is used for solving stiff systems of ordinary differential equations. Also a procedure for training neural networks based on the method of back propagation of error is employed together with the Adam optimization algorithm, that is a modified version of the gradient descent supplied with an automatic step adjustment. Results. It is shown that the neural network mappings built according to the described procedure are able to reproduce the dynamics of single neural oscillators. Moreover, without additional training, these mappings can be used as elements of a coupled system for the dynamics modeling of coupled neural oscillator systems. Conclusion. The described neural network mapping can be considered as a new universal framework for complex dynamics modeling. In contrast to models based on series expansion (power, trigonometric), neural network mapping does not require truncating of the series. Consequently, it allows modeling processes with arbitrary order of nonlinearity, hence there are reasons to believe that in some aspects it will be more effective. The approach developed in this paper based on the neural network mapping can be considered as a sort of an alternative to the traditional numerical methods of modeling of dynamics. What makes this approach topical is the current rapid development of technologies for creating fast computing equipment that supports neural network training and operation.
- 10 December 2023
- Published by Saratov State University in Izvestiya VUZ. Applied Nonlinear Dynamics
- Vol. 32 (1) , 72-95
Open AccessModeling of the Hodgkin–Huxley neural oscillators dynamics using an artificial neural network
The purpose of this study — to represent a detailed description of the procedure for creating and training a neural network mapping on the example of the dynamics modeling of a neural oscillator of the Hodgkin–Huxley type; to show that the neural network mappings trained for single oscillators can be used as elements of a coupled system that simulate the behavior of coupled oscillators. Methods. Numerical method is used for solving stiff systems of ordinary differential equations. Also a procedure for training neural networks based on the method of back propagation of error is employed together with the Adam optimization algorithm, that is a modified version of the gradient descent supplied with an automatic step adjustment. Results. It is shown that the neural network mappings built according to the described procedure are able to reproduce the dynamics of single neural oscillators. Moreover, without additional training, these mappings can be used as elements of a coupled system for the dynamics modeling of coupled neural oscillator systems. Conclusion. The described neural network mapping can be considered as a new universal framework for complex dynamics modeling. In contrast to models based on series expansion (power, trigonometric), neural network mapping does not require truncating of the series. Consequently, it allows modeling processes with arbitrary order of nonlinearity, hence there are reasons to believe that in some aspects it will be more effective. The approach developed in this paper based on the neural network mapping can be considered as a sort of an alternative to the traditional numerical methods of modeling of dynamics. What makes this approach topical is the current rapid development of technologies for creating fast computing equipment that supports neural network training and operation.
- 27 January 2023
- Published by IOP Publishing in Laser Physics
- Vol. 33 (2) , 025004
Modelling light wave propagation and interaction and dynamical regimes analysis using the Lyapunov exponents method in ring fibre cavities
The ‘Cabaret’ finite differences method allows the performance of fast and conclusive modelling of the microcavity temporal dynamics with group velocity dispersion, cross- and self- phase modulation included. The proposed scheme and model allow us to study the dynamics of a cavity with two counter-propagating pulses, second order dispersion, modulation instability, Rayleigh scattering and other effects along the cavity roundtrips number up to 106…7. The possibility of Lyapunov exponents mapping for the Raman and SBS fibre lasers and cavities is discussed.
- 27 January 2023
- Published by IOP Publishing in Laser Physics
- Vol. 33 (2) , 025004
Modelling light wave propagation and interaction and dynamical regimes analysis using the Lyapunov exponents method in ring fibre cavities
The ‘Cabaret’ finite differences method allows the performance of fast and conclusive modelling of the microcavity temporal dynamics with group velocity dispersion, cross- and self- phase modulation included. The proposed scheme and model allow us to study the dynamics of a cavity with two counter-propagating pulses, second order dispersion, modulation instability, Rayleigh scattering and other effects along the cavity roundtrips number up to 106…7. The possibility of Lyapunov exponents mapping for the Raman and SBS fibre lasers and cavities is discussed.
- 21 December 2022
- Published by Elsevier in Chaos, Solitons & Fractals
- Vol. 167
Discovering dynamical features of Hodgkin–Huxley-type model of physiological neuron using artificial neural network
- 21 December 2022
- Published by Elsevier in Chaos, Solitons & Fractals
- Vol. 167
Discovering dynamical features of Hodgkin–Huxley-type model of physiological neuron using artificial neural network
- 1 November 2022
- Published by Pleiades Publishing Ltd in Regular and Chaotic Dynamics
- Vol. 27 (6) , 697-712
Synchronization and Bistability of Two Uniaxial Spin-Transfer Oscillators with Field Coupling
A spin-transfer oscillator is a nanoscale device demonstrating self-sustained precession of its magnetization vector whose length is preserved. Thus, the phase space of this dynamical system is limited by a three-dimensional sphere. A generic oscillator is described by the Landau – Lifshitz – Gilbert – Slonczewski equation, and we consider a particular case of uniaxial symmetry when the equation yet experimentally relevant is reduced to a dramatically simple form. The established regime of a single oscillator is a purely sinusoidal limit cycle coinciding with a circle of sphere latitude (assuming that points where the symmetry axis passes through the sphere are the poles). On the limit cycle the governing equations become linear in two oscillating magnetization vector components orthogonal to the axis, while the third one along the axis remains constant. In this paper we analyze how this effective linearity manifests itself when two such oscillators are mutually coupled via their magnetic fields. Using the phase approximation approach, we reveal that the system can exhibit bistability between synchronized and nonsynchronized oscillations. For the synchronized one the Adler equation is derived, and the estimates for the boundaries of the bistability area are obtained. The two-dimensional slices of the basins of attraction of the two coexisting solutions are considered. They are found to be embedded in each other, forming a series of parallel stripes. Charts of regimes and charts of Lyapunov exponents are computed numerically. Due to the effective linearity the overall structure of the charts is very simple; no higher-order synchronization tongues except the main one are observed.
- 1 November 2022
- Published by Pleiades Publishing Ltd in Regular and Chaotic Dynamics
- Vol. 27 (6) , 697-712
Synchronization and Bistability of Two Uniaxial Spin-Transfer Oscillators with Field Coupling
A spin-transfer oscillator is a nanoscale device demonstrating self-sustained precession of its magnetization vector whose length is preserved. Thus, the phase space of this dynamical system is limited by a three-dimensional sphere. A generic oscillator is described by the Landau – Lifshitz – Gilbert – Slonczewski equation, and we consider a particular case of uniaxial symmetry when the equation yet experimentally relevant is reduced to a dramatically simple form. The established regime of a single oscillator is a purely sinusoidal limit cycle coinciding with a circle of sphere latitude (assuming that points where the symmetry axis passes through the sphere are the poles). On the limit cycle the governing equations become linear in two oscillating magnetization vector components orthogonal to the axis, while the third one along the axis remains constant. In this paper we analyze how this effective linearity manifests itself when two such oscillators are mutually coupled via their magnetic fields. Using the phase approximation approach, we reveal that the system can exhibit bistability between synchronized and nonsynchronized oscillations. For the synchronized one the Adler equation is derived, and the estimates for the boundaries of the bistability area are obtained. The two-dimensional slices of the basins of attraction of the two coexisting solutions are considered. They are found to be embedded in each other, forming a series of parallel stripes. Charts of regimes and charts of Lyapunov exponents are computed numerically. Due to the effective linearity the overall structure of the charts is very simple; no higher-order synchronization tongues except the main one are observed.