Search

In this paper, we survey selected software packages for the numerical solution of boundary value ODEs (BVODEs), time-dependent PDEs in one spatial dimension (1DPDEs), and initial value ODEs (IVODEs). A unifying theme of this paper is our focus on software packages for these problem classes that compute error-controlled, continuous numerical solutions. A continuous numerical solution can be accessed by the user at any point in the domain. We focus on error-control software; this means that the software adapts the computation until it obtains a continuous approximate solution with a corresponding error estimate that satisfies the user tolerance. The second section of the paper will provide an overview of recent work on the development of COLNEWSC, an updated version of the widely used collocation BVODE solver, COLNEW, that returns an error-controlled continuous approximate solution based on the use of a superconvergent interpolant to the underlying collocation solution. The third section of the paper gives a brief review of recent work on the development of a new 1DPDE solver, BACOLIKR, that provides time- and space-dependent event detection for an error-controlled continuous numerical solution. In the fourth section of the paper, we briefly review the state of the art in IVODE software for the computation of error-controlled continuous numerical solutions. [ABSTRACT FROM AUTHOR]

This paper deals with the geometric numerical integration of gradient flow and its application to optimization. Gradient flows often appear as model equations of various physical phenomena, and their dissipation laws are essential. Therefore, dissipative numerical methods, which are numerical methods replicating the dissipation law, have been studied in the literature. Recently, Cheng, Liu, and Shen proposed a novel dissipative method, the Lagrange multiplier approach, for gradient flows, which is computationally cheaper than existing dissipative methods. Although their efficacy is numerically confirmed in existing studies, the existence results of the Lagrange multiplier approach are not known in the literature. In this paper, we establish some existence results. We prove the existence of the solution under a relatively mild assumption. In addition, by restricting ourselves to a special case, we show some existence and uniqueness results with concrete bounds. As gradient flows also appear in optimization, we further apply the latter results to optimization problems. [ABSTRACT FROM AUTHOR]

This paper proposes a framework for constructing high-order linearly implicit exponential integrators that conserve a quadratic invariant. This is then applied to the scalar auxiliary variable (SAV) approach. Quadratic invariants are significant objects that are present in various physical equations and also in computationally efficient conservative schemes for general invariants. For instance, the SAV approach converts the invariant into a quadratic form by introducing scalar auxiliary variables, which have been intensively studied in recent years. In this vein, Sato et al. (Appl. Numer. Math. 187, 71-88 2023) proposed high-order linearly implicit schemes that conserve a quadratic invariant. In this study, it is shown that their method can be effectively merged with the Lawson transformation, a technique commonly utilized in the construction of exponential integrators. It is also demonstrated that combining the constructed exponential integrators and the SAV approach yields schemes that are computationally less expensive. Specifically, the main part of the computational cost is the product of several matrix exponentials and vectors, which are parallelizable. Moreover, we conduct some mathematical analyses on the proposed schemes. [ABSTRACT FROM AUTHOR]

In this paper, the approximate solution of a set of nonlinear third order differential equations with mixed boundary conditions is obtained by employing the fifth step block method and Modified Taylor Series Scheme (MTSS). The motivation of this subject to implement MTSS for determining the shooting constants associated with the Initial Value Problems (IVPs) rather than Boundary Value Problems (BVPs). The fifth step block method is also used to solve nonlinear third Order Differential Equations (ODEs) on the definite interval. Two numerical examples are experimented to demonstrate the efficiency and accuracy of the proposed scheme by obtaining the absolute errors. Furhter, the order, convergence and stability of the proposed method are discussed to strengthen the theoretical concept. [ABSTRACT FROM AUTHOR]

In this paper we consider the numerical solution of fractional differential equations. In particular, we study a step-by-step procedure, defined over a graded mesh, which is based on a truncated expansion of the vector field along the orthonormal Jacobi polynomial basis. Under mild hypotheses, the proposed procedure is capable of getting spectral accuracy. A few numerical examples are reported to confirm the theoretical findings. [ABSTRACT FROM AUTHOR]

In this paper, we present a reformulation of Mickens' rules for nonstandard finite difference (NSFD) schemes to adapt them to systems of ODEs. We want to compare several methods within the class of NSFD schemes. These methods are expressed as exponential integrators which treat exactly the linear part of the equations. While the classical methods address each equation of a system separately, we propose here to address them as a whole, which involves a reformulation of Mickens' rules. This leads to enhance the results compared to the classical case, even if approximations of the matrix exponentials are used. Precisely, using Cayley–Hamilton theorem we obtain two correction factors, one of which is linked to the size of the system and the other to the nonlinear term, we are thus able to use test cases which show the impact of one or another corrector. We prove the consistency and the zero-stability of the method in the nonlinear case to conclude to the convergence of the method. [ABSTRACT FROM AUTHOR]

Runge–Kutta embedded pairs of high algebraic order are frequently utilized when strict tolerances are required. When creating such pairings of orders nine and eight for use in double precision arithmetic, numerous conditions are often satisfied. First and foremost, we strive to keep the coefficients' magnitudes small to prevent accuracy loss. We may, however, allow greater coefficients when working with quadruple precision. Then, we may build pairs of orders 9 and 8 with significantly smaller truncation errors. In this paper, a novel pair is generated that, as predicted, outperforms state-of-the-art pairs of the same orders in a collection of important problems. [ABSTRACT FROM AUTHOR]

This paper considers the regularization continuation method and the trust-region updating strategy for the nonlinearly equality-constrained optimization problem. Namely, it uses the inverse of the regularization quasi-Newton matrix as the pre-conditioner and the Sherman–Morrison–Woodbury formula to improve its computational efficiency in the well-posed phase, and it adopts the inverse of the regularization two-sided projection of the Hessian as the pre-conditioner to improve its robustness in the ill-conditioned phase. Since it only solves a linear system of equations and the Sherman–Morrison–Woodbury formula significantly saves the computational time at every iteration, and the sequential quadratic programming (SQP) method needs to solve a quadratic programming subproblem at every iteration, it is faster than SQP. Numerical results also show that it is more robust and faster than SQP (the built-in subroutine fmincon.m of the MATLAB2020a environment and the subroutine SNOPT executed in GAMS v28.2 (GAMS Corporation, 2019) environment). The computational time of the new method is about one third of that of fmincon.m for the large-scale problem. Finally, the global convergence analysis of the new method is also given. [ABSTRACT FROM AUTHOR]

Structure-preserving algorithms and algorithms with uniform error bound have constituted two interesting classes of numerical methods. In this paper, we blend these two kinds of methods for solving nonlinear systems with highly oscillatory solution, and the blended algorithms inherit and respect the advantage of each method. Two kinds of algorithms are presented to preserve the symplecticity and energy of the Hamiltonian systems, respectively. Long time energy conservation is analysed for symplectic algorithms and the proposed algorithms are shown to have uniform error bound in the position for the highly oscillatory structure. Moreover, some methods with uniform error bound in the position and in the velocity are derived and analysed. Two numerical experiments are carried out to support all the theoretical results established in this paper by showing the performance of the blended algorithms. [ABSTRACT FROM AUTHOR]

The main aim of the current paper is to propose a new numerical scheme based on the Adams-Bashforth method for solving two component time fractional differential equations involving the Caputo-Prabhakar fractional derivative of order μ. The existence and uniqueness of the solutions based on the Adams-Bashforth method reported for two component differential equations are discussed. Also, in this paper stability and convergence are demonstrated. Moreover, two numerical example are demonstrated in order to show the validity and reliability of the proposed methods. [ABSTRACT FROM AUTHOR]

The paper presents a novel method that allows one to establish numerical solutions of linear and nonlinear ordinary differential equations—with polynomial coefficients—that contain any finite products of the unknown functions and/or their general derivatives. The presented algorithm provides numerical solutions of these differential equations subject to initial or boundary conditions. This algorithm proposes the desired solution in terms of B-polynomials (Bernstein polynomial basis) and then uses the orthonormal relation of B-polynomials with its weighted dual basis with respect to the Jacobi weight function to construct a linear/nonlinear system in the unknown expansion coefficients which can be solved using a suitable solver. The properties of B-polynomials provide greater flexibility in which to impose the initial or boundary conditions at the end points of the interval [0, R] and enable us to obtain exactly and explicitly some of the unknown expansion coefficients in the form of a suggested numerical solution. Consequently, the presented algorithm leads to a linear or nonlinear algebraic system in the unknown expansion coefficients that has a simpler form than that was obtained by the other algorithms. So that, this procedure is a powerful tool that we may utilize to overcome the difficulties associated with boundary and initial value problems with less computational effort than the other techniques. An accepted agreement is obtained between the exact and approximate solutions for the given examples. The error analysis was also studied, and the obtained numerical results clarified the validity of the theoretical results. [ABSTRACT FROM AUTHOR]

In this paper, we consider a two-point boundary-value problem with a Riemann–Liouville–Caputo fractional derivative of order α ∈ (1 , 2) . We solve this boundary-value problem in a sequence of processes, first using the shooting technique based on the secant iterative method, we convert the boundary-value problem into an initial-value problem, then the initial-value problem is transformed into an equivalent Volterra integral equation with weakly singular kernel. Finally, we find the approximate solution of the resultant equation by using a discretization scheme on a uniform mesh. The method's convergence analysis has been thoroughly established, and it demonstrates that the scheme is first-order convergent. To show the effectiveness of the suggested approach, numerical results are provided. [ABSTRACT FROM AUTHOR]

This paper reviews strong stability preserving discrete variable methods for differential systems. The strong stability preserving Runge–Kutta methods have been usually investigated in the literature on the subject, using the so-called Shu–Osher representation of these methods, as a convex combination of first-order steps by forward Euler method. In this paper, we revisit the analysis of strong stability preserving Runge–Kutta methods by reformulating these methods as a subclass of general linear methods for ordinary differential equations, and then using a characterization of monotone general linear methods, which was derived by Spijker in his seminal paper (SIAM J Numer Anal 45:1226–1245, 2007). Using this new approach, explicit and implicit strong stability preserving Runge–Kutta methods up to the order four are derived. These methods are equivalent to explicit and implicit RK methods obtained using Shu–Osher or generalized Shu–Osher representation. We also investigate strong stability preserving linear multistep methods using again monotonicity theory of Spijker. [ABSTRACT FROM AUTHOR]

This paper investigates numerical solution of generalized space-time fractional Klein–Gordon equations (GSTFKGE) by using Gegenbauer wavelet method (GWM). The developed method makes use of fractional order integral operator (FOIO) for Gegenbauer wavelet, which is constructed by employing the definition of Riemann–Liouville fractional integral (RLFI) operator and Laplace transformation. The present algorithm is based on Gegenbauer wavelet jointly with FOIO to convert a GSTFKGE into a system of equations which is solved by using Newton's technique. Additionally, the upper bound of error norm of the proposed method is calculated to validate the theoretical authenticity of the developed method. The comparison of numerical outcomes with the existing results in the literature and graphical illustrations show the accuracy and reliability of our method. [ABSTRACT FROM AUTHOR]

In this paper we introduce a procedure for identifying optimal methods in parametric families of numerical schemes for initial value problems in partial differential equations. The procedure maximizes accuracy by adaptively computing optimal parameters that minimize a defect-based estimate of the local error at each time step. Viable refinements are proposed to reduce the computational overheads involved in the solution of the optimization problem, and to maintain conservation properties of the original methods. We apply the new strategy to recently introduced families of conservative schemes for the Korteweg-de Vries equation and for a nonlinear heat equation. Numerical tests demonstrate the improved efficiency of the new technique in comparison with existing methods. [ABSTRACT FROM AUTHOR]

In this paper, we present a novel investigation of the so-called SAV approach, which is a framework to construct linearly implicit geometric numerical integrators for partial differential equations with variational structure. SAV approach was originally proposed for the gradient flows that have lower-bounded nonlinear potentials such as the Allen–Cahn and Cahn–Hilliard equations, and this assumption on the energy was essential. In this paper, we propose a novel approach to address gradient flows with unbounded energy functionals such as the KdV equation by a decomposition of energy functionals. Further, we will show that the equation of the SAV approach, which is a system of equations with scalar auxiliary variables, is expressed as another gradient system that inherits the variational structure of the original system. This expression allows us to construct novel higher-order integrators by a certain class of Runge-Kutta methods. We will propose second and fourth order schemes for conservative systems in our framework and present several numerical examples. [ABSTRACT FROM AUTHOR]

To integrate large systems of nonlinear differential equations in time, we consider a variant of nonlinear waveform relaxation (also known as dynamic iteration or Picard–Lindelöf iteration), where at each iteration a linear inhomogeneous system of differential equations has to be solved. This is done by the exponential block Krylov subspace (EBK) method. Thus, we have an inner-outer iterative method, where iterative approximations are determined over a certain time interval, with no time stepping involved. This approach has recently been shown to be efficient as a time-parallel integrator within the PARAEXP framework. In this paper, convergence behavior of this method is assessed theoretically and practically. We examine efficiency of the method by testing it on nonlinear Burgers, Liouville–Bratu–Gelfand, and nonlinear heat conduction equations and comparing its performance with that of conventional time-stepping integrators. [ABSTRACT FROM AUTHOR]

This paper develops an improved tri-coloured rooted-tree theory for the order conditions for ERKN methods solving general multi-frequency and multidimensional second-order oscillatory systems. The bottleneck of the original tricoloured rooted-tree theory is the existence of numerous redundant trees. In light of the fact that the sum of the products of the symmetries and the elementary differentials is meaningful, this paper naturally introduces the so-called extended elementary differential mappings. Then, the new improved tri-coloured rooted tree theory is established based on a subset of the original tri-coloured rooted-tree set. This new theory makes all redundant trees disappear, and thus, the order conditions of ERKN methods for general multi-frequency and multidimensional second-order oscillatory systems are reduced greatly. Furthermore, with this new theory, we present some new ERKN methods of order up to four. Numerical experiments are implemented and the results show that ERKN methods can be competitive with other existing methods in the scientific literature, especially when comparatively large stepsizes are used. [ABSTRACT FROM AUTHOR]

Recently, the efficient numerical solution of Hamiltonian problems has been tackled by defining the class of energy-conserving Runge-Kutta methods named Hamiltonian Boundary Value Methods (HBVMs). Their derivation relies on the expansion of the vector field along the Legendre orthonormal basis. Interestingly, this approach can be extended to cope with other orthonormal bases and, in particular, we here consider the case of the Chebyshev polynomial basis. The corresponding Runge-Kutta methods were previously obtained by Costabile and Napoli (Numer. Algo., 27, 119–130 2021). In this paper, the use of a different framework allows us to carry out a novel analysis of the methods also when they are used as spectral formulae in time, along with some generalizations of the methods. [ABSTRACT FROM AUTHOR]

In this paper, we discuss the numerical solution of systems of differential equations with both linear (stiff) and nonlinear components that arise from the semi-discretization of certain partial differential equations. When faced with the task of solving problems with discontinuous solutions numerically, the linear stability properties are insufficient for convergence. So, to overcome this drawback, strong stability preserving (SSP) methods are introduced. While using SSP implicit-explicit methods lead to severe constraints on the allowed time-stepping, an integrating factor approach reduces this limitation by solving the linear part exactly. In this work, we develop integrating factor general linear methods (IFGLMs) with strong stability properties. Sufficient conditions for IFGLMs to be SSP are discussed. The construction of IFGLMs of order p ≤ 6 and stage order 1 ≤ q ≤ p with r = p external stages and 2 ≤ s ≤ 10 internal stages is presented, which have larger effective SSP coefficients than the class of Runge–Kutta and two–step Runge–Kutta methods. Results are verified numerically on several representative test cases. [ABSTRACT FROM AUTHOR]

Getting an efficient neural network can be a very difficult task for engineers and researchers because of the huge number of hyperparameters to tune and their interconnections. To make the tuning step easier and more understandable, this work focuses on probably one of the most important leverage to improve Neural Networks efficiency: the optimizer. These recent years, a great number of algorithms have been developed but they need an accurate tuning to be efficient. To get rid of this long and experimental step, we are looking for generic and desirable properties for non-convex optimization. For this purpose, the optimizers are reinterpreted or analyzed as a discretization of a continuous dynamical system. This continuous framework offers many mathematical tools in order to interpret the sensitivity of the optimizer with respect to the initial guess such as Lyapunov stability. By enforcing the discrete decrease of Lyapunov functionals, new robust and efficient optimizers are designed. They also considerably simplify the tuning of hyperparameters (learning rate, momentum etc.). These Lyapunov based algorithms outperform several state of the art optimizers on different benchmarks of the literature. Drawing its inspiration from the numerical analysis of PDEs, this paper emphasizes the essential role of some hidden energy/entropy quantities for machine learning tasks. [ABSTRACT FROM AUTHOR]

The paper exhibits a practical and effective scheme to approximate the solutions of a class of fractional differential equations known as the Bagley-Torvik equations. The underlying fractional derivative is based on the Caputo definition. Both the boundary and initial conditions are considered while the domain of approximation is taken sufficiently large. The principle idea behind our approximation algorithm is to use a novel class of (orthogonal) polynomials called the Bessel polynomials, which their coefficients are positive. The convergence and error analysis of the Bessel solution are investigated in the L 2 and infinity norms. Representing the unknown solution and its derivatives in terms of basis functions together with collocation points, the Bagley-Torvik equations are reduced into an algebraic form. In particular, we present a simple but fast algorithm with linear complexity for computing the q-th order fractional derivative of the basis functions in the vectorized form. Several practical test problems are given to illustrate the utility and efficiency of the proposed approximation algorithm. Validation of the method is obtained by comparison with existing available numerical solutions. Based on experiments, the present approach produces the results of high accuracy to obtain the approximate solutions of Bagley-Torvik equations especially for long-time computations. [ABSTRACT FROM AUTHOR]

In this paper, we consider model order reduction for bilinear systems with non-zero initial conditions. We discuss the choices of Gramians for both the homogeneous and the inhomogeneous parts of the system individually and prove how these Gramians characterise the respective dominant subspaces of each of the two subsystems. Proposing different, not necessarily structure preserving, reduced-order methods for each subsystem, we establish several strategies to reduce the dimension of the full system. For all these approaches, error bounds are shown depending on the truncated Hankel singular values of the subsystems. Besides the error analysis, stability is discussed. In particular, a focus is on a new criterion for the homogeneous subsystem guaranteeing the existence of the associated Gramians and an asymptotically stable realisation of the system. [ABSTRACT FROM AUTHOR]

In this paper, the system of nonlinear inequalities is considered. The problem is approximated by the parameterized smooth equations which is formed by Chen-Harker-Kanzow-Smale smoothing function. A smoothing homotopy method is proposed to solve the system of smooth equations. For almost all initial points in Euclidean space, the existence and convergence of the homotopy path is proved under certain weaker condition without assumption of the monotonicity. The numerical results illustrate that the method is effective. [ABSTRACT FROM AUTHOR]

In last few years, many ERKN methods have been investigated for solving multi-frequency multidimensional second-order ordinary differential equations, and the numerical efficiency has been checked strongly in scientific computation. But in the constructions of (especially high-order) new ERKN methods, lots of time and effort are costed in presenting the practical order conditions firstly and then in adding some reasonable assumptions to get the coefficient functions finally. In this paper, a feasible and effective technique is given which makes the construction of ERKN methods finished in a few seconds or a few minutes, even for high-order integrators. Moreover, this technique does not need any more information and knowledge except the classical RKN method. And this paper also gives the theoretical explanation to guarantee that the ERKN method obtained from this technique has the same order and the same properties as the underlying RKN method. [ABSTRACT FROM AUTHOR]

Many chemical systems exhibit a range of patterns, a noticeable and interesting class of numerical patterns that arise in autocatalytic reactions which changes with increasing spatial domains. In this paper, autocatalytic spatiotemporal patterns were demonstrated using the system of chemical species modeled with the time-fractional Caputo derivatives of subdiffusive orders. It is not new that a spectral algorithm with its entire nature is more accurate when compared with a finite-difference scheme to solve a range of integer and non-integer order time partial differential equations. This is because the Fourier spectral techniques have the upper hand on high-order spectral accuracy, and are computationally efficient. Hence, it is regarded as the best approach to existing lower-order methods for integrating the second-order partial derivatives in space. This motivates the present study to explore the usefulness of Fourier spectral methods in resolving and obtaining complex Turing patterns arising from nonlinear fractional autocatalytic reaction-diffusion problems in high dimensions. The autocatalysis model was examines for linear stability in an attempt to obtain the correct choice of parameters that are likely to lead to the formation of new complex turing-like patterns. Numerical experiments in the 2D lead to a striking range of patterns arising from catalytic reactions of fractional-order labyrinthine pattern-like structures. Analysis of pattern formation was also extended to 3D dynamics to obtain a new set of patterns like a star-, cyclic-, diamond-like, and the emergence of apple-shaped structures, which are greatly influenced by either the choice parameters that are involved or that of the initial conditions. [ABSTRACT FROM AUTHOR]

In this paper, a split-step θ algorithm is proposed for nonlinear stochastic variable delay differential equations. It is proved that the numerical solution obtained by the algorithm converges to the exact solution under the local Lipschitz condition. Moreover, it is shown that the numerical algorithm can maintain the mean square stability of the analytical solution for linear scalar equations. Finally, numerical experiments demonstrate the correctness of the theoretical algorithm. [ABSTRACT FROM AUTHOR]

This paper presents a parareal algorithm with parameterized propagators for linear parametric differential equations over a wide range of parameters. Through transforming the initial value problem into nonparametric ODEs based on Taylor series, we construct the general parameterized fine and coarse propagators for the parareal algorithm in each time subinterval. Furthermore, to accelerate the convergence of the algorithm, the coarse propagator based on the waveform relaxation method is proposed. By analysing the computational complexity of the WR propagator and the general coarse propagator, we find these two propagators are appropriate for the different situations. Finally, the convergence analysis of the parareal algorithm with these propagators is presented and our analysis is illustrated with two numerical experiments. [ABSTRACT FROM AUTHOR]

We show that the probability of the exceptional set decays exponentially for a broad class of randomized algorithms approximating solutions of ODEs, admitting a certain error decomposition. This class includes randomized explicit and implicit Euler schemes, and the randomized two-stage Runge-Kutta scheme (under inexact information). We design a confidence interval for the exact solution of an IVP and perform numerical experiments to illustrate the theoretical results. [ABSTRACT FROM AUTHOR]

In this paper, we construct and analyze a new dynamical low-rank integrator for second-order matrix differential equations. The method is based on a combination of the projector-splitting integrator introduced in Lubich and Oseledets (BIT 54(1):171–188, 2014. ) and a Strang splitting. We also present a variant of the new integrator which is tailored to semilinear second-order problems. [ABSTRACT FROM AUTHOR]

Dynamical low-rank integrators for matrix differential equations recently attracted a lot of attention and have proven to be very efficient in various applications. In this paper, we propose a novel strategy for choosing the rank of the projector-splitting integrator of Lubich and Oseledets adaptively. It is based on a combination of error estimators for the local time-discretization error and for the low-rank error with the aim to balance both. This ensures that the convergence of the underlying time integrator is preserved. The adaptive algorithm works for projector-splitting integrator methods for first-order matrix differential equations and also for dynamical low-rank integrators for second-order equations, which use the projector-splitting integrator method in its substeps. Numerical experiments illustrate the performance of the new integrators. [ABSTRACT FROM AUTHOR]