Kolmogorov Equation Research Articles

Variable linear transformation improved physics-informed neural networks to solve thin-layer flow problems

Physics-informed neural networks (PINNs) have attracted wide attention due to their ability to seamlessly embed the learning process with physical laws and their considerable success in solving forward and inverse differential equation (DE) problems. While most studies are improving the learning process and network architecture of PINNs, less attention has been paid to the modification of the DE system, which may play an important role in addressing some limitations of PINNs. One of the simplest modifications that can be implemented to all DE systems is the variable linear transformation (VLT). Therefore, in this work, we propose the VLT-PINNs that solve the DE systems of the linear-transformed variables instead of the original ones. To clearly illustrate the importance of prior knowledge in determining the VLT parameters, we choose the thin-layer flow problems as our focus. Ten related cases were tested, including the jet flows, wake flows, mixing layers, boundary layers and Kovasznay flows. Based on the principle of normalization and for a better match of the DE system to the preference of NNs, we identify three principles for determining the VLT parameters: magnitude normalization for dependent variables (principle 1), local normalization for independent variables (principle 2), and appropriate scaling for physics-related parameters in inverse problems (principle 3). The VLT-PINNs with the VLT parameters suggested by the proposed principles show excellent performance over all the test cases, while the results are quite poor with the VLT parameters suggested by traditional linear transformations, such as nondimensionalization and global normalization. Comparison studies also show that only under the constraints of the VLT principles can we obtain satisfactory results. Besides, we find tanh is more appropriate as the activation function than sin for thin-layer flow problems, from both posteriori results and priori analyses with physical intuition. We highlight that our VLT method is an attempt to combine the three advantages of accuracy, universality and simplicity, and hope that it can provide new insights into the better integration of prior knowledge, physical intuition and the nature of NNs. The code for this paper is available on https://github.com/CAME-THU/VLT-PINN.

A space-time second-order method based on modified two-grid algorithm with second-order backward difference formula for the extended Fisher–Kolmogorov equation

In this paper, a modified two-grid algorithm based on block-centred finite difference method is developed for the fourth-order nonlinear extended Fisher–Kolmogorov equation. To further improve the computational efficiency, an effective second-order accurate backward difference formula is considered. The modified two-grid method based on Newton iteration is constructed to linearize the nonlinear system. The method solves a miniature nonlinear system on a coarse grid accompanying a larger time step to get the numerical solution, then computes a linear system constructed by the previous result with the Taylor expansion on a fine grid accompanying a smaller time step to get the correct numerical solution. Theoretical analysis shows that the modified two-grid algorithm can achieve second-order convergence accuracy both in time and space domain. Several numerical experiments are provided to verify the theoretical result and the high efficiency of this approach. The practical problem illustrates the actual applicable value of the algorithm.

Fully analytic approach to study bifurcation phenomenon for a shape memory alloy beam under temperature variation and lateral white-noise excitation

The main aim of this article is to obtain the probability density function of the response of an SMA beam under lateral white-noise excitation and study the bifurcation phenomenon through a fully analytic approach. Firstly, an efficient constitutive stress–strain relation for a shape memory alloy material supporting both martensite and austenite phases is considered. Secondly, the governing equation of motion for a typical SMA simply supported beam is derived using some dimensionless parameters. Thirdly, the corresponding probability density function of the response is computed analytically using the powerful Fokker–Planck–Kolmogorov equation, and finally, the bifurcation phenomenon for parameter variations of the beam is investigated. The results show how geometric (beam aspect ratio), force (the mean value of the white-noise excitation), and environmental (working temperature) factors can affect the nonlinear behavior and response bifurcation of the beam. A numerical validation is also done that guarantees the correctness of the method and results.

DriftRec: Adapting Diffusion Models to Blind JPEG Restoration.

In this work, we utilize the high-fidelity generation abilities of diffusion models to solve blind JPEG restoration at high compression levels. We propose an elegant modification of the forward stochastic differential equation of diffusion models to adapt them to this restoration task and name our method DriftRec. Comparing DriftRec against an L2 regression baseline with the same network architecture and state-of-the-art techniques for JPEG restoration, we show that our approach can escape the tendency of other methods to generate blurry images, and recovers the distribution of clean images significantly more faithfully. For this, only a dataset of clean/corrupted image pairs and no knowledge about the corruption operation is required, enabling wider applicability to other restoration tasks. In contrast to other conditional and unconditional diffusion models, we utilize the idea that the distributions of clean and corrupted images are much closer to each other than each is to the usual Gaussian prior of the reverse process in diffusion models. Our approach therefore requires only low levels of added noise and needs comparatively few sampling steps even without further optimizations. We show that DriftRec naturally generalizes to realistic and difficult scenarios such as unaligned double JPEG compression and blind restoration of JPEGs found online, without having encountered such examples during training.

Duality between controllability and observability of stochastic linear evolution equations with general filtration

This article is devoted to establishing a general duality for controllability and observability of forward and backward stochastic linear evolution equations with general filtration. The main tool that we employ is the stochastic transposition method introduced in [13]. First, we prove the well-posedness of controlled forward and backward stochastic linear evolution equations with unbounded control operators. Then, we obtain the equivalence between controllability of stochastic evolution equations and observability for their dual equations. Finally, an example is given for illustrating our general results.

Stochastic modeling of spreading an infection with standard incidence rate

This paper studies and models the random spread of an infection in a population. It extends the traditional deterministic modeling approach by incorporating discretetime stochastic modeling using Markov chains. The probability of extinction and disease persistence is then investigated using the branching chain method, with a focus on a quantity in the non-random model, which is called the basic reproductive number. Moreover, the random model is transformed into a system of stochastic differential equations by approximating the probability distribution function using the forward Kolmogorov equation. An equivalent stochastic differential system is also introduced and examined for the stochastic model. Finally, numerical simulations of the stochastic models are conducted to evaluate the theoretical findings presented in the paper.

Combining closed-form and numerical solutions for the inverse kinematics of six-degrees-of-freedom collaborative handling robot

In the process of solving the inverse kinematics of six-degrees-of-freedom collaborative robots, the numerical solution has problems such as low accuracy and singular configurations. Moreover, due to the high coupling of its position and attitude, the direct closed-form solution fails. To address these problems, an inverse kinematics algorithm that combines closed-form and numerical solutions was proposed. The Jacobian matrix was established based on the forward kinematics equation of the six-degrees-of-freedom collaborative robot. Its inverse matrix was obtained by a singular value decomposition of the matrix using the Manocha elimination method to avoid the singularities of the Jacobian matrix. The optimal inverse kinematics solution was obtained using the Newton–Raphson iterative method. A computer simulation implemented in MATLAB and Visual C++ was used to evaluate the accuracy and speed of the proposed algorithm. The results of the trajectory planning experiments demonstrate that the algorithm satisfies the requirements for real-time smooth motion. This study of inverse kinematics lays the theoretical foundation for solutions to the inverse problem and real-time control of collaborative robots.

Khas'minskii's Theorem for the Kolmogorov Equation with Partially Singular Diffusion Matrix

Исследуется стационарное уравнение Колмогорова с вырожденной матрицей диффузии и разрывным коэффициентом сноса. Получены достаточные условия существования вероятностного решения. Построены примеры, показывающие точность условий. Библиография: 15 названий.

Anomalous diffusion limit for a kinetic equation with a thermostatted interface

We consider the limit of solutions of scaled linear kinetic equations with a reflection-transmission-killing condition at the interface. Both the coefficient describing the probability of killing and the scattering kernel degenerate. We prove that the long-time, large-space limit is the unique solution of a version of the fractional in space heat equation that corresponds to the Kolmogorov equation for a symmetric stable process, which is reflected, or transmitted while crossing the interface and is killed upon the first hitting of the interface. The results of the paper are related to the work in Komorowski et al. (Ann Prob 48:2290–2322, 2020), where the case of a non-degenerate probability of killing has been considered.

Optimal Weak Order and Approximation of the Invariant Measure with a Fully-Discrete Euler Scheme for Semilinear Stochastic Parabolic Equations with Additive Noise

In this paper, we consider the ergodic semilinear stochastic partial differential equation driven by additive noise and the long-time behavior of its full discretization realized by a spectral Galerkin method in spatial direction and an Euler scheme in the temporal direction, which admits a unique invariant probability measure. Under the condition that the nonlinearity is once differentiable, the optimal convergence orders of the numerical invariant measures are obtained based on the time-independent weak error, but not relying on the associated Kolmogorov equation. More precisely, the obtained convergence orders are O(λN−γ) in space and O(τγ) in time, where γ∈(0,1] from the assumption ∥Aγ−12Q12∥L2 is used to characterize the spatial correlation of the noise process. Finally, numerical examples confirm the theoretical findings.

Physics-informed machine learning method with space-time Karhunen-Loève expansions for forward and inverse partial differential equations

We propose a physics-informed machine-learning method based on space-time-dependent Karhunen-Loève expansions (KLEs) of the state variables and the residual least-square formulation of the solution of partial differential equations. This method, which we name dPICKLE, results in a reduced-order model for solving forward and inverse time-dependent partial differential equations. By conditioning KLEs on data, dPICKLE seamlessly assimilates data in forward and inverse solutions. KLEs are linear in unknown parameters. Because of this, and unlike physics-informed deep-learning methods based on the residual least-square formulation, for well-posed partial differential equation (PDE) problems, dPICKLE leads to linear least-square problems (directly for linear PDEs and after linearization for nonlinear PDEs), which guarantees a unique solution. The efficiency and accuracy of dPICKLE are demonstrated for linear and nonlinear forward and inverse problems via comparison with analytical, finite difference, and physics-informed neural network (PINN) solutions.

Non-stationary probabilistic analysis of non-linear ship roll motion due to modulated periodic and random excitations

This paper is about the study on the non-stationary probability density function (PDF) solution of the non-linear ship rolling system under modulated periodic and random excitations. The quadratic and cubic damping terms are introduced to describe the non-linear damping moment and the power terms in state variables are adopted to describe the righting moment in the ship rolling system, respectively. The relevant Fokker–Planck–Kolmogorov (FPK) equation of the ship rolling system is constructed and the non-stationary probabilistic solutions are obtained by the presented solution procedure. In the solution process, the time variable t is introduced to the traditional exponential-polynomial-closure (EPC) method to obtain the non-stationary PDF solutions of the ship rolling system. The obtained non-stationary PDF solutions of the non-linear ship rolling systems demonstrate that the presented solution procedure can give high solution accuracy even in the PDF tails when they are compared with Monte Carlo simulation (MCS). Moreover, owing to the strong system nonlinearity and the interaction of the periodic excitation and random excitation, the achieved non-stationary PDF solutions of the system responses can deviate significantly from the Gaussian PDF. It is also found that the PDFs of responses are not even symmetrical about the means of responses due to the interaction of the periodic excitation and the random excitation even if the system only contains odd nonlinear terms. In addition, the computational efficiency is increased by over five hundred times by the presented solution procedure compared to MCS.

A candidate method for prediction of the non-stationary response of strongly nonlinear systems under wide-band noise excitation

This paper expands the application of the RBFNN method to the prediction of non-stationary responses in strongly nonlinear systems subjected to wide-band noises (WBNs) excitation. Specifically, the SAM is initially employed to deduce the averaged Fokker–Planck–Kolmogorov (FPK) equation. The non-stationary responses of the energy of the system under WBNs are modeled as a one-dimensional Markov diffusion process. This yields a one-dimensional FPK equation that governs the NSRs of the system subject to WBNs. Subsequently, solving the corresponding FPK equation and obtaining the system’s NSRs are accomplished with the RBFNN method. Furthermore, four indicative cases exposed to WBNs are studied to verify the proposed scheme. Comparisons with simulations demonstrate the accuracy and efficiency of the developed technique.

Propagation of random excitations in the polar representation of strongly nonlinear four-wave systems of turbulence.

The present work is concerned with the uncertainty propagation of the wave turbulent system. In particular, we study the temporal development and long-term behavior of the probability with respect to the amplitude and phase of complex-valued waves constituting the generic four-wave system of turbulence. Our approach to approximating the target distribution function is via the three steps: (i) to grasp the physical process described by the true turbulence model as random process, (ii) to determine the stochastic differential equation whose solution exhibits statistically similar behavior with the underlying turbulent signal, and (iii) to solve the corresponding Kolmogorov forward equation. Our implementation of the methodology is distinguished by employing a number of simplified stochastic models and applying one of them in the adaptive fashion which varies subject to the different parameter regime of the true dynamical system model. Accordingly, we become able to demonstrate the effectiveness of this reduced-order modeling framework for the analysis of the turbulent system characterized by not only weak but strong interactions among the nonlinear waves. We numerically corroborate our theoretical predictions in the context of the generalized Majda-Mclaughlin-Tabak wave turbulence prototype.

Application of the Conjugate Gradient Method for Predicting Unknown Heat Flux in 2D Plat under Effect of Insulation

When an anomaly is discovered during a thermal inspection of industrial equipment, the diagnosis of finding the heat source is made more difficult by the presence of heat insulation. The use of the Conjugate Gradient Method (CGM) to solve the unknown heat flux in Inverse Heat Conduction Problem (IHCP) under insulation condition has been discussed. This method is used to predict a heat flux by utilizing temperature data at any points of measurement. The study deals with the structure of 2D thin plate at steady state. A copper plate with a thickness of 1 mm is used as specimen. White glass wool is used to thermally insulate the plate on the back surface. An electrical heater is plate mounted and operated by varying the DC voltage. A datalogger with thermocouple sensors records plate temperature changes. Software for solving the IHCP based on the CGM was developed. Initially, the software was implemented using synthetic data obtained from the forward heat conduction equations with a known heat flux. Simulated measurement data are assumed to have high accuracy (± 0.1 ºC). Determination of heat flux with the CGM gives very satisfactory results with a low error below 0.12%. The CGM was then applied using experimental data. With a measurement error of ± 0.5 ºC, the computational stopping criterion can still be met satisfactorily. In real installations, equipment is sometimes partially or completely isolated. In the experiment, this phenomenon is simulated by varying the isolation area as follows: 0%, 25%, 50% and 100%. The results of the flux calculation are correlated with the real flux supplied by the electrical heater. Each type of insulation area has a similar curve. By obtaining the correlation equation, the equivalent heat flux from calculation to real and vice versa can be carried out.

Necessary and sufficient optimal control conditions for networked control systems with input delay and Markovian packet dropouts

AbstractThis article studies the optimal control for networked control systems (NCSs) subject to Markovian packet dropouts and input delay. In this scenario, Markovian packet dropouts occur in two channels, that is, from sensor to controller and from controller to actuator. Meanwhile, the input delay arises in the channel from controller to actuator. The contributions of this paper are as the following two aspects: Firstly, based on Pontryagin's maximum principle, the solution to the forward and backward stochastic difference equations (FBSDEs) is obtained. Secondly, in virtue of the obtained solution to the FBSDEs, sufficient and necessary optimal control conditions are derived and the explicit expression of the optimal controller is presented. The validity of the results is illustrated by numerical examples.

The stochastic stability and H∞‐fuzzy control of stochastic bifurcation of a doubly‐fed induction generator

AbstractConsidering the dynamic behavior of doubly‐fed induction generators (DFIGs) under the influence of random factors, this paper not only analyzes the phenomenon of stochastic instability and bifurcation of a DFIG dynamic variable in its random space when they are affected by environmental noise, but also proposes a method based on the Tkagi–Sugneo (T–S) fuzzy control strategy to control its stochastic bifurcation. First, a four‐dimensional stochastic dynamic DFIG model is established by using multiplicative white noise to simulate the influence of environmental noise on electrical variables, and stochastic central manifold theory is used to reduce the dimensionality of a planar model in the bifurcation neighborhood. Then, the stochastic stability of the model is investigated based on singular boundary theory, while the steady‐state probability density of the stochastic DFIG is determined using the Fokker–Plank–Kolmogorov equation to obtain the location and probability density of its stochastic P‐bifurcation. Finally, the influence of stochastic bifurcation behavior is eliminated by H∞‐fuzzy output feedback control. The numerical simulation results indicate that the location and probability of stochastic bifurcation in a DFIG will vary with the change in noise intensity, and the bifurcation parameter values and stochastic stability domain are obtained. The harm caused by random factors can be solved based on H∞‐fuzzy output feedback, which provides a theoretical basis for the stable operation of the DFIG system.

Dimensional Optimization of a Modular Robot Manipulator

The mechanism parameters of the manipulator not only have a great influence on the size of the working space but also affect flexible performance distribution. Aimed at obtaining a 6 DOF modular manipulator, mechanism parameters were optimized in order to explore the effect of upper arm and forearm dimensions on the end dexterity of the manipulator. First, forward kinematic equations were derived using the DH method, and the Jacobian matrix of the manipulator was solved. Second, three indicators, including the condition number index, structural length index, and global conditioning index, were employed as optimization indicators for the mechanism parameters of the manipulator, and an orthogonal experiment was designed based on the Grey–Taguchi method and robot toolbox. Third, the grey relational analysis method was used to process the experimental results, and the grey relational grade for each group was solved. Last, the variation curve between the grey relational grade and the parameter level of each mechanism was drawn, and optimized mechanical arm mechanism parameters were derived. It was found that although the overall dimension of the manipulator was slightly decreased, as determined via comparing the original and optimized manipulator length, the performance indexes were improved. The results not only verified the correctness of the proposed optimization method but also laid a foundation for subsequent research on the dynamic performance of modular robot systems.

Decentralized Output-Feedback Control With Asymmetric One-Step Delayed Information

This paper considers the decentralized LQG control problem for discrete-time decentralized system controlled by two players. In this scenario, player 1 shares a unit delayed observations and control inputs with the controller of player 2, whereas due to the limiting capacity, the controller of player 1 cannot obtain the observations and control inputs of player 2 which leads to the asymmetric one-step delay information. It should be emphasized that this structure makes the classical separation principle fail. Under the assumption of linear control strategies, we derive the optimal estimators based on asymmetric information. In virtue of the Pontryagin's maximum principle, the optimal explicit decentralized controllers are obtained by solving the forward and backward stochastic difference equations (FBSDEs). It is noted that the control gains are coupled with the estimation gains. Moreover, the estimation gains satisfy the forward Riccati equations and the control gains follow the backward Riccati equations which cannot be solved simultaneously. To this end, we present iterative solutions to the coupled forward and backward Riccati equations. Finally, a sufficient condition for the stabilization problem is given in terms of the coupled algebraic Riccati equations. Numerical examples are illustrated to show the effectiveness of the proposed algorithm.

Superinfection and the hypnozoite reservoir for Plasmodium vivax: a general framework

A characteristic of malaria in all its forms is the potential for superinfection (that is, multiple concurrent blood-stage infections). An additional characteristic of Plasmodium vivax malaria is a reservoir of latent parasites (hypnozoites) within the host liver, which activate to cause (blood-stage) relapses. Here, we present a model of hypnozoite accrual and superinfection for P. vivax. To couple host and vector dynamics for a homogeneously-mixing population, we construct a density-dependent Markov population process with countably many types, for which disease extinction is shown to occur almost surely. We also establish a functional law of large numbers, taking the form of an infinite-dimensional system of ordinary differential equations that can also be recovered by coupling expected host and vector dynamics (i.e. a hybrid approximation) or through a standard compartment modelling approach. Recognising that the subset of these equations that model the infection status of the human hosts has precisely the same form as the Kolmogorov forward equations for a Markovian network of infinite server queues with an inhomogeneous batch arrival process, we use physical insight into the evolution of the latter process to write down a time-dependent multivariate generating function for the solution. We use this characterisation to collapse the infinite-compartment model into a single integrodifferential equation (IDE) governing the intensity of mosquito-to-human transmission. Through a steady state analysis, we recover a threshold phenomenon for this IDE in terms of a parameter R_0 expressible in terms of the primitives of the model, with the disease-free equilibrium shown to be uniformly asymptotically stable if R_0<1 and an endemic equilibrium solution emerging if R_0>1. Our work provides a theoretical basis to explore the epidemiology of P. vivax, and introduces a strategy for constructing tractable population-level models of malarial superinfection that can be generalised to allow for greater biological realism in a number of directions.