Modified Extended Lie Group Method for Hessenberg Differential Algebraic Equations with Index 3

Preprint

Article

Modified Extended Lie Group Method for Hessenberg Differential Algebraic Equations with Index 3

Altmetrics

Downloads

102

Views

Comments

A peer-reviewed article of this preprint also exists.

This version is not peer-reviewed

Submitted:

14 April 2023

Posted:

14 April 2023

You are already at the latest version

Alerts

Abstract

Hessenberg differential algebraic equations (Hessenberg-DAEs) with high index play a critical role in the modeling of mechanical systems and multibody dynamics. Motivated by the widely used Lie Group Differential Algebraic Equation (LGDAE) method which only handles index 2 systems, we propose a Modified Extended Lie Group Differential Algebraic Equation (MELGDAE) method for solving index 3 Hessenberg-DAEs, and provide theoretical analysis to deepen the foundation of the MELGDAE method.The performance of the MELGDAE method is compared with the standard methods RADAU and MEBDF on index 2 and 3 DAE systems, and it is demonstrated that the MELGDAE integrator exhibits competitive performance in terms of high accuracy and the preservation of algebraic constraints. In particular, all differential variables in index 3 Hessenberg DAEs achieve second-order convergence using the MELGDAE method, which suggests potential for extension to Hessenberg-DAEs with an index of 4 or higher.

Keywords:

Subject: Computer Science and Mathematics - Applied Mathematics

1. Introduction

Differential-Algebraic Equations (DAEs) are a class of mathematical models that arise in various fields, such as multibody dynamics [1], electrical circuits [2], fluid mechanics [3], optimal control [4] and chemical reactions [5]. DAEs are a natural extension of ordinary differential equations that also involve algebraic equations. As such, they offer a convenient way to model complex systems that involve both dynamic and algebraic components. The study of DAEs has been an active area of research for many years, and numerous numerical methods have been developed to solve these DAEs, particularly for low index systems. These methods include well known Runge-Kutta methods [6,7,8], backward differentiation formulae [9,10],and others. However, the efficient solution of DAEs remains a challenging problem due to the presence of algebraic equations that may be nonlinear, implicit, and of high index.

Hessenberg differential algebraic equations (Hessenberg-DAEs) are an important type of differential equation which are also commonly used in the modeling of mechanical systems and electrical circuits.Because of their unique structure, numerical methods for solving Hessenberg-DAEs can differ significantly from those used for other types of differential equations. Over the years, some approaches have been proposed for the numerical solution of Hessenberg-DAEs, including variational iteration method [11], adomian decomposition method for [12], exponential integrators [13]. These methods either lack rigorous theoretical analysis or are only suitable for low-index systems.

One promising approach is the Lie group method whose several variations have been developed specifically for solving (Hessenberg) DAEs with high index in recent years. Lie group methods mostly constitute one-step extensions of classical methods, including Runge-Kutta type formulas [14]. In mechanical engineering, BDF methods, which have traditionally been used, have been extended to Lie groups as well [15]. Despite their usefulness, these methods suffer from a numerical disadvantage. Various numerical techniques can be used to compute the matrix exponential, but the properties of these methods, such as stability, accuracy, and efficiency, can impact the results of the Lie-group methods. Based on the structural information of Hessenberg DAEs, C.S.Liu [16,17,18] propose a Lie-group differential algebraic equations (LGDAE) method. By converting the canonical representation of the governing equations for nonholonomic systems, the LGDAE method expresses them in the form of DAE systems, comprising of nonlinear ordinary differential equations and nonlinear algebraic equations. The method is composed of two components: the first component pertains to the Lie-group structure and concerns the numerical solution of nonlinear ordinary differential equations. The second component, the Newton iterative scheme, is used to solve the nonlinear algebraic equations. The LGDAE method has been successfully employed to solve various practical problems such as the heat source recovery problem [19,20], sliding mode control problem [21], and inverse nonlinear vibration problems [22]. However, it is noteworthy that the LGDAE method is only applicable to systems with index 2.

The main focus of this paper is to investigate Lie group methods for Hessenberg-DAEs with an index of 3 or higher. We present a modified extended Lie group method that incorporates the benefits of Liu’s original method, along with the structural characteristics of Hessenberg-DAEs. Our proposed method enables the accurate solution of high-index Hessenberg-DAEs and effectively captures the behavior of solutions to algebraic constraints. We demonstrate the effectiveness of our method through numerical experiments on some benchmark problems.

The structure of this paper is as follows. Section 2 presents a brief introduction to Hessenberg-DAEs and Lie group methods. In Section 3, we present our novel modified extended Lie group DAE method for Hessenberg-DAEs with an index of 3,and provide a theoretical justification for the proposed method. Section 4 presents the results of numerical experiments on several benchmark problems. Finally, we conclude with a discussion of our results and directions for future research in Section 5.

2. Preliminaries

In this section, we provide a concise overview of Hessenberg-DAEs and an implicit

G L (n, R)

Lie group methods.

2.1. Hessenberg DAEs

A system of ordinary differential equations (ODEs) with algebraic constraints which can be written in the general form as

f (t, x (t), x^{'} (t)) = 0

(1)

is called differential algebraic equation, where

\frac{\partial f}{\partial x}

may be singular and the structure and rank of the of this Jacobian matrix may depend on the solution

x (t)

. An important special case of 1 is the semi-explicit DAE of the form

\{\begin{matrix} {x_{1}}^{'} (t) & = & f_{1} (t, x_{1} (t), x_{2} (t)) \\ 0 & = & f_{2} (t, x_{1} (t), x_{2} (t)), \end{matrix}

(2)

whose index is 1 if the Jacobian matrix function

\frac{\partial f_{2}}{\partial x_{2}}

is assumed to be nonsingular for all t. Hessenberg Index-1 is also often referred to as a semi-explicit index-1 system.

Definition 1.

Hessenberg Index-2

\{\begin{matrix} {x_{1}}^{'} (t) & = & f_{1} (t, x_{1} (t), x_{2} (t)), (3 a) \\ 0 & = & f_{2} (t, x_{1} (t)), (3 b) \end{matrix}

where

f_{i}

are sufficiently differentiable, the variables

x_{2}

of the algebraic part is absent from the constraint and the product of Jacobians

\frac{\partial f_{2}}{\partial x_{1}} \frac{\partial f_{1}}{\partial x_{2}}

is nonsingular. Thus, all algebraic variables play the role of index-2 therefore this is a pure index-2 DAE.

Definition 2.

Hessenberg Index-3

\{\begin{matrix} {x_{1}}^{'} (t) & = & f_{1} (t, x_{1} (t), x_{2} (t), x_{3} (t)), (4 a) \\ {x_{2}}^{'} (t) & = & f_{2} (t, x_{1} (t), x_{2} (t)), (4 b) \\ 0 & = & f_{3} (t, x_{2} (t)) . (4 c) \end{matrix}

In this case the product of three matrix functions

F = \frac{\partial f_{3}}{\partial x_{2}} \frac{\partial f_{2}}{\partial x_{1}} \frac{\partial f_{1}}{\partial x_{3}}

is nonsingular, where

x_{1}

x_{2}

and

x_{3}

are

n_{1}

n_{2}

and

n_{3}

dimensions respectively. The index of a Hessenberg DAE is found, as in the general case, by differentiation. However, only the algebraic constraints must be differentiated.

Definition 3.

Hessenberg Index-r

\{\begin{matrix} {x_{1}}^{'} & = & f_{1} (t, x_{1}, x_{2}, \dots, x_{r}), (5 a) \\ {x_{2}}^{'} & = & f_{2} (t, x_{1}, x_{2}, \dots, x_{r - 1}), (5 b) \\ ⋮ & (5 c) \\ {x_{i}}^{'} & = & f_{i} (t, x_{i - 1}, x_{i}, \dots, x_{r - 1}), 3 \leq i \leq r - 1, (5 d) \\ ⋮ & (5 e) \\ 0 & = & f_{r} (t, x_{r - 1}) . (5 f) \end{matrix}

In this case the product of three matrix functions

\frac{\partial f_{r}}{\partial x_{r - 1}} \frac{\partial f_{r - 1}}{\partial x_{r - 2}} \dots \frac{\partial f_{2}}{\partial x_{1}} \frac{\partial f_{1}}{\partial x_{r}}

is nonsingular.

2.2. Lie group structure of ODEs

A Lie group is a differentiable manifold equipped with a group structure that is consistent with the underlying topology of the manifold. The general linear group is a Lie group, whose manifold is an open subset

G L (n, R) : = {G \in R^{n \times n} | det G \neq 0}

. Thus,

G L (n, R)

is also an

n \times n

-dimensional manifold. The group composition is given by the matrix multiplication.

The general linear group

G L (n, R)

gives uniquely a real Lie-algebra

g l (n, R)

. Consider a one-parameter subgroup

G (t)

t \in R

, of the general linear group

G L (n, R)

, which is a curve passing through the group identity at

t = 0

G (0) = I_{n}

, and which left acts on the n-dimensional Euclidean space

R^{n}

, resulting in a congruence of curves in

R^{n}

X (t) = G (t) X (0) = [G (t) G^{- 1} (t_{0})] X (t_{0}), t_{0}, t \in R .

(6)

Owing to the closure property of the Lie group,

G (t) G^{- 1} (t_{0})

also belongs to

G L (n, R)

. When

t_{0}

is put very close to t,

G (t) G^{- 1} (t_{0})

is very close to the identity

I_{n}

. Moreover,

A (t) : = \frac{\partial}{\partial t} [G (t) G^{- 1} (t_{0})] |_{t_{0} = t} = \dot{G} (t) G^{- 1} (t)

(7)

defines a string of tangent vectors on the tangent space at the group identity of the group manifold, more precisely, a continuously singly parametrized series of one-dimensional sub-algebra of the real Lie algebra

g l (n, R)

Differentiating Eq. (6), setting

t_{0} = t

and then using Eq. (7) yields

X {(t)}^{'} = A (t) X (t) .

(8)

The flow generated by such a

g l (n, R)

vector field is the congruence of curves resulting from solving the system (8). Essentially, it is very important that when one wants to establish a local coordinate on a smooth manifold on which a finite dimensional Lie-group of transformations is acting, the solution of the differential equations system is necessary.

The Lie-group method possesses a greater advantage than other numerical methods due to its Lie-group structure, and it is a powerful technique to solve ordinary differential equations (ODEs). Here we give a new form of the dynamics in Eq. (4a) or (4b) from the

G L (n, R)

Lie-group structure. In order to fit the format in Eq. (8), the vector field

f_{i}

on the right-hand side of Eq. (4a) or (4b) can be written as

\{\begin{matrix} x_{1}^{'} & = & A_{1} (t, x_{1}, x_{2}, x_{3}) x_{1} (9 a) \\ x_{2}^{'} & = & A_{2} (t, x_{1}, x_{2}) x_{2} (9 b) \end{matrix}

where

A_{i} = \frac{f_{i}}{| | x_{i} | |} \otimes \frac{x_{i}}{| | x_{i} | |}

is the coefficient matrix and

a \otimes b

denotes the dyadic operation of

a

and b, i.e.,

(a \otimes b) c = (b \cdot c) a

. Here we suppose that

| | x_{i} | | > 0

2.3. An implicit $G L (n, R)$ Lie-group method

The equation (9) provides a new starting point for constructing the Lie-group

G L (n, R)

approach. In order to develop a numerical scheme, we suppose that the coefficient matrix

A_{i}

is constant with

{\bar{a}}_{i} = \frac{{\bar{f}}_{i}}{| | {\bar{x}}_{i} | |}, {\bar{b}}_{i} = \frac{{\bar{x}}_{i}}{| | {\bar{x}}_{i} | |},

(10)

which can be obtained by taking the values of

f_{i}

and

x_{i}

at a suitable mid-point of

\bar{t} \in [t_{0} = 0, t]

, where

t \leq t_{0} + h

and h is a small time stepsize.

By introducing new variables

w_{i} (t) = {\bar{b}}_{i}^{T} \cdot x_{i} (t)

, we obtain an augmented dynamical system

\frac{d}{d t} (\begin{matrix} x_{i} (t) \\ w_{i} (t) \end{matrix}) = (\begin{matrix} 0 & {\bar{a}}_{i} \\ 0 & {\bar{a}}_{i}^{T} {\bar{b}}_{i} \end{matrix}) (\begin{matrix} x_{i} (t) \\ w_{i} (t) \end{matrix}) .

(11)

Because the state matrix in Eq. (11) can be decomposed as

(\begin{matrix} 0 & {\bar{a}}_{i} \\ 0 & {\bar{a}}_{i}^{T} {\bar{b}}_{i} \end{matrix}) = (\begin{matrix} I_{n} & \frac{{\bar{a}}_{i}}{{\bar{a}}_{i}^{T} {\bar{b}}_{i}} \\ 0 & 1 \end{matrix}) (\begin{matrix} I_{n} & 0 \\ 0 & {\bar{a}}_{i}^{T} {\bar{b}}_{i} \end{matrix}) {(\begin{matrix} I_{n} & \frac{{\bar{a}}_{i}}{{\bar{a}}_{i}^{T} {\bar{b}}_{i}} \\ 0 & 1 \end{matrix})}^{- 1},

(12)

the solution of the linear dynamical system can be written analytically as follows

(\begin{matrix} x_{i} (t) \\ w_{i} (t) \end{matrix}) = exp [(\begin{matrix} 0 & {\bar{a}}_{i} \\ 0 & {\bar{a}}_{i}^{T} {\bar{b}}_{i} \end{matrix}) (t - t_{k})] (\begin{matrix} x_{i} (t_{k}) \\ w_{i} (t_{k}) \end{matrix}) = (\begin{matrix} I_{n} & ρ (c_{i}, t - t_{k}) {\bar{a}}_{i} \\ 0 & exp [{\bar{c}}_{i} (t - t_{k})] \end{matrix}) (\begin{matrix} x_{i} (t_{k}) \\ w_{i} (t_{k}) \end{matrix}),

(13)

where

ρ (u, t) = \frac{e^{u t} - 1}{u}

c_{i} = {\bar{a}}_{i}^{T} {\bar{b}}_{i}

. We reduce the augmented

x_{i} - w_{i}

system to the

x_{i}

system, obtaining

x_{i} (t) = [I_{n} + ρ (c_{i}, t - t_{k}) {\bar{a}}_{i} {\bar{b}}_{i}^{T}] x_{i} (t_{k}) = G_{i} (t - t_{k}) x_{i} (t_{k}),

(14)

and

det G_{i} (t - t_{k}) > 0

. Note that

G_{i}

is indeed an element in the Lie group

G L (n, R)

. Therefore, by introducing the auxiliary variable

w_{i}

and the augmented system, we are able to derive an explicit expression for the components of a one-parameter group

G L (n, R)

3. MELGDAE method for index-3 Hessenberg-DAEs

In this section, we present a novel modified extended Lie group DAE (MELGDAE) method for Hessenberg-DAEs with an index of 3,and provide a theoretical justification for the proposed method.

We first calculate the solution

x_{2} (t_{k + 1})

by the given variables

x_{1} (t_{k})

x_{2} (t_{k})

, then obtain the solution

x_{1} (t_{k + 1})

via given variables

x_{1} (t_{k})

x_{3} (t_{k})

and

x_{2} (t_{k + 1})

. Moreover, we compute the solution

x_{3} (t_{k + 1})

by the resulting value of

x_{1} (t_{k + 1})

and

x_{2} (t_{k + 1})

. Hereafter, we denote the loop that solves for the variable

x_{i}

as the

x_{i}

-loop. The MELGDAE method, which is presented in detail below, utilizes the notation

x_{i [k]} = x_{i} (t_{k})

: For the $x_{2}$ -Loop:
S1:: Give $0 \leq θ_{2} \leq 1$ , $x_{1 [k]}$ , $x_{2 [k]}$ .
S2:: Estimate ${\bar{x}}_{2 [k + 1]}$ via Euler method as follow,

${\bar{x}}_{2 [k + 1]} = x_{2 [k]} + h f_{2} (t_{k}, x_{1 [k]}, x_{2 [k]}) .$

(15)
S3:: Let $τ_{2} = t_{k} + θ_{2} h$ .
S4:: Compute $\hat{x_{2 [k + 1]}}$ :

S5:: If ${\hat{x}}_{2 [k + 1]}$ convergers according to a given stopping criterion

$| | {\hat{x}}_{2 [k + 1]} - {\bar{x}}_{2 [k + 1]} | | < ε_{2},$

(17)

then let $x_{2 [k + 1]} = {\bar{x}}_{2 [k + 1]}$ ; otherwise, let ${\bar{x}}_{2 [k + 1]} = {\hat{x}}_{2 [k + 1]}$ and goto S4.

: For the $x_{1}$ -Loop:
S1:: Give $0 \leq θ_{1} \leq 1$ , $x_{1 [k]}$ , $x_{3 [k]}$ , $x_{2 [k + 1]}$ resulting from the $x_{2}$ -Loop.
S2:: Estimate ${\bar{x}}_{1 [k + 1]}$ via Euler method as follow,

${\bar{x}}_{1 [k + 1]} = x_{1 [k]} + h f_{1} (t_{k}, x_{1 [k]}, x_{2 [k + 1]}, x_{3 [k]}) .$

(18)
S3:: Let $τ_{1} = t_{k} + θ_{1} h$ .
S4:: Compute $\hat{x_{1 [k + 1]}}$ :

S5:: If ${\hat{x}}_{1 [k + 1]}$ convergers according to a given stopping criterion

$| | {\hat{x}}_{1 [k + 1]} - {\bar{x}}_{1 [k + 1]} | | < ε_{1},$

(20)

then let $x_{1 [k + 1]} = {\bar{x}}_{1 [k + 1]}$ ; otherwise, let ${\bar{x}}_{1 [k + 1]} = {\hat{x}}_{1 [k + 1]}$ and goto S4.

: For the $x_{3}$ -Loop:
S1:: Let $τ_{3} = t_{k} + θ_{3} h$ , ${\bar{x}}_{3 [k + 1]} = x_{3 [k]}$ .
S2:: Compute ${\hat{x}}_{3 [k + 1]}$ via Newton iterative method:

${\hat{x}}_{3 [k + 1]} = {\bar{x}}_{3 [k + 1]} - J^{- 1} f_{3} (t_{k + 1}, x_{2} ([x_{1} ({\bar{x}}_{3 [k + 1]})])),$

(21)

where ${\tilde{x}}_{1 [k]} = (1 - θ_{3}) x_{1 [k]} + θ_{3} x_{1 [k + 1]}$ , ${\tilde{x}}_{2 [k]} = (1 - θ_{3}) x_{2 [k]} + θ_{3} x_{2 [k + 1]}$ , $J = \frac{\partial f_{3}}{\partial x_{3}} = \frac{\partial f_{3}}{\partial x_{2}} \frac{\partial x_{2}}{\partial x_{1}} \frac{\partial x_{1}}{\partial x_{3}} |_{x_{1} = {\tilde{x}}_{1 [k]}, x_{2} = {\tilde{x}}_{2 [k]}, x_{3} = {\bar{x}}_{3 [k + 1]}}$ . We could obtain $\frac{\partial x_{1}}{\partial x_{3}} |_{x_{1} = {\tilde{x}}_{1 [k]}, x_{2} = {\tilde{x}}_{2 [k]}, x_{3} = {\bar{x}}_{3 [k + 1]}}$ , $\frac{\partial x_{2}}{\partial x_{1}} |_{x_{1} = {\tilde{x}}_{1 [k]}, x_{2} = {\tilde{x}}_{2 [k]}}$ from Eq. (14).

S3:: If ${\hat{x}}_{3 [k + 1]}$ convergers according to a given stopping criterion

$| | {\hat{x}}_{3 [k + 1]} - {\bar{x}}_{3 [k + 1]} | | < ε_{3},$

(24)

then let $x_{3 [k + 1]} = {\bar{x}}_{3 [k + 1]}$ ; otherwise, let

${\bar{x}}_{3 [k + 1]} = {\hat{x}}_{3 [k + 1]},$

$x_{2 [k + 1]} = {\tilde{x}}_{2 [k + 1]},$

$x_{1 [k + 1]} = {\tilde{x}}_{1 [k + 1]},$

and goto S2.

Lemma 1.

([23]) Let

U

be an

n \times m

matrix and

V

be an

m \times n

matrix. Then

λ^{m} | λ I_{n} + U V | = λ^{n} | λ I_{m} + V U |,

(25)

where λ is any scalar and

I_{k}

denotes the

k \times k

identity matrix.

Lemma 2.

Let

U

be an

n \times m

matrix,

V

be an

m \times n

matrix, and

a

and

b

be non-zero

m \times 1

vectors. Define the

m \times m

matrix

E = I_{m} + h a b^{T}

for

h > 0

. If

U V

is nonsingular and

h < \frac{1}{{| | Q | |}_{2} \cdot {| | a | |}_{2} \cdot {| | b | |}_{2}}

, where

Q = V {(U V)}^{- 1} U

, then

U

V

is nonsingular.

Proof.

To prove that

U E V

is nonsingular, we need to show that its determinant is nonzero.

To begin with,

U V

is nonsingluar, that is,

n < = m

. And it is noted that

Q^{2} = Q

, then

{| | Q | |}_{2} > = 1

For case

n = m

, we obtain

\begin{matrix} det (U E V) & = det (U) det (E) det (V) \\ = det (U) det (V) det (E) \\ = det (U V) det (E) \end{matrix}

and using Lemma 1, we have

det (E) = 1 + h b^{T} a > 0,

where

h < \frac{1}{{| | Q | |}_{2} \cdot {| | a | |}_{2} \cdot {| | b | |}_{2}}

Since both

det (U V)

and

det (E)

are nonzero, then

det (U E V)

is also nonzero. Therefore

U E V

is nonsingular.

For case

n < m

, using Lemma 1, we obtain

\begin{matrix} det (U E V) & = det (U V + h U a b^{T} V) \end{matrix}

(26)

\begin{matrix} = det (U V) det (I_{n} + h {(U V)}^{- 1} U a b^{T} V) \end{matrix}

(27)

\begin{matrix} = det (U V) (1 + h b^{T} V {(U V)}^{- 1} U a) \end{matrix}

(28)

Moreover,

| b^{T} {Q a | \leq | | b | |}_{2} \cdot {| | Q a | |}_{2} \leq {| | Q | |}_{2} \cdot {| | a | |}_{2} \cdot {| | b | |}_{2}

, then

1 + h b^{T} Q a > 0

, which means that

U E V

is nonsingular from equation (28). □

Lemma 3.

Let

U_{1}

be an

n_{1} \times n_{3}

matrix,

U_{2}

be an

n_{2} \times n_{1}

matrix, and

U_{3}

be an

n_{3} \times n_{2}

matrix. Let

a_{1}

and

b_{1}

be nonzero

n_{1} \times 1

vectors, and let

a_{2}

and

b_{2}

be nonzero

n_{2} \times 1

vectors. Define the matrices

E_{1} = I_{n_{1}} + h a_{1} b_{1}^{T}

and

E_{2} = I_{n_{2}} + h a_{2} b_{2}^{T}

, where

h > 0

. Suppose that

U_{3} U_{2} U_{1}

is nonsingular and that h satisfies

h < min {\frac{1}{3 \cdot | | Q_{1} {| |}_{2} \cdot | | a_{1} {| |}_{2} \cdot | | b_{1} {| |}_{2}}, \frac{1}{3 \cdot | | Q_{2} {| |}_{2} \cdot | | a_{2} {| |}_{2} \cdot | | b_{2} {| |}_{2}}, \frac{1}{{2 | | M | |}_{2}}},

where

Q_{1} = U_{1} {(U_{3} U_{2} U_{1})}^{- 1} U_{3} U_{2}

Q_{2} = U_{2} U_{1} {(U_{3} U_{2} U_{1})}^{- 1} U_{3}

, and

M = U_{1} {(U_{3} U_{2} U_{1})}^{- 1} U_{3}

a_{2} b_{2}^{T} U_{2}

. Then, the matrix

U_{3} E_{2} U_{2} E_{1} U_{1}

is nonsingular.

Proof.

Firstly,

U_{3} U_{2} U_{1}

is nonsingular, that is

n_{3} \leq n_{1}

and

n_{3} \leq n_{2}

Let

V_{2} = U_{2} U_{1},

that is

U_{3} V_{2}

is nonsingular. And

E_{2} = I_{n_{2}} + h a_{2} b_{2}^{T}

and

h < \frac{1}{3 | | Q_{2} {| |}_{2} \cdot | | a_{2} {| |}_{2} \cdot | | b_{2} {| |}_{2}}

, where

Q_{2} = U_{2} U_{1} {(U_{3} U_{2} U_{1})}^{- 1} U_{3}

. By Lemma 2,

U_{3} E_{2} V_{2}

is nonsingular, which means

U_{3} E_{2} U_{2} U_{1}

is nonsingular.

Then, let

U^{'} = U_{3} E_{2} U_{2},

and

U^{'} U_{1}

is nonsingular. In addtion,

E_{1} = I_{n_{1}} + h a_{1} b_{1}^{T}

, if

h < \frac{1}{| | Q_{1}^{'} {| |}_{2} \cdot | | a_{1} {| |}_{2} \cdot | | b_{1} {| |}_{2}}

, where

Q_{1}^{'} = U_{1} {(U_{3} E_{2} U_{2} U_{1})}^{- 1} U_{3} E_{2} U_{2}

. By Lemma 2,

U^{'} E_{1} U_{1}

is nonsingular, which means

U_{3} E_{2} U_{2} E_{1} U_{1}

is nonsingular.

Now, we need proof

h | | Q_{1}^{'} {| |}_{2} \cdot | | a_{1} {| |}_{2} \cdot | | b_{1} {| |}_{2} < 1 .

(29)

It is noted that

Q_{1}^{'} = {(I_{n_{1}} + h M)}^{- 1} Q_{1} (I_{n_{1}} + h M),

(30)

where

h < \frac{1}{{2 | | M | |}_{2}}

. Moreover, we are able to obtain

| | {(I_{n_{1}} + h M)}^{- 1} {| |}_{2} \leq \frac{1}{1 - {| | h M | |}_{2}} < 2

and

| | I_{n_{1}} + {h M | |}_{2} \leq | | I_{n_{1}} {| |}_{2} + {| | h M | |}_{2} < \frac{3}{2}

Then, we obtain

\begin{matrix} | | Q_{1}^{'} {| |}_{2} & = | | {(I_{n_{1}} + h M)}^{- 1} Q_{1} (I_{n_{1}} + h M) {| |}_{2} \end{matrix}

(31)

\begin{matrix} \leq | | {(I_{n_{1}} + h M)}^{- 1} {| |}_{2} \cdot | | Q_{1} {| |}_{2} \cdot | | I_{n_{1}} + h M {| |}_{2} \end{matrix}

(32)

\begin{matrix} < 3 | | Q_{1} {| |}_{2} \end{matrix}

(33)

For

h < \frac{1}{3 | | Q_{1} {| |}_{2} \cdot | | a_{1} {| |}_{2} \cdot | | b_{1} {| |}_{2}}

, then we obtain the inequality (29). □

Theorem 1.

For the

x_{3}

-Loop of Hessenberg DAEs with index 3, if h is sufficiently small, then the Jacobian matrix

J

is nonsingular.

Proof.

We first compute the

\frac{\partial x_{2}}{\partial x_{1}}

by Eq.(22).

\begin{matrix} \frac{\partial x_{2}}{\partial x_{1}} & = {\tilde{d}}_{12 [k]} {{\tilde{a}}_{12 [k]} \frac{\partial}{\partial x_{1}} {\tilde{ρ}}_{2 [k]} + {\tilde{ρ}}_{2 [k]} \frac{\partial}{\partial x_{1}} {\tilde{a}}_{12 [k]}} \end{matrix}

(34)

\begin{matrix} = {\tilde{d}}_{12 [k]} {{\tilde{a}}_{12 [k]} [\frac{(u t - 1) exp (u t) + 1}{u^{2}} |_{u = {\tilde{c}}_{12 [k]}, t = h}] \frac{\partial}{\partial x_{1}} {\tilde{c}}_{12 [k]} + {\tilde{ρ}}_{2 [k]} \frac{\partial}{\partial x_{1}} {\tilde{a}}_{12 [k]}} \end{matrix}

(35)

\begin{matrix} = {\tilde{d}}_{12 [k]} {{\tilde{a}}_{12 [k]} \frac{({\tilde{c}}_{12 [k]} h - 1) exp ({\tilde{c}}_{12 [k]} h) + 1}{{\tilde{c}}_{12 [k]}^{2}} [{\tilde{b}}_{12 [k]}^{T} \frac{\partial}{\partial x_{1}} {\tilde{a}}_{12 [k]}] + {\tilde{ρ}}_{2 [k]} \frac{\partial}{\partial x_{1}} {\tilde{a}}_{12 [k]}} \end{matrix}

(36)

\begin{matrix} = {\tilde{d}}_{12 [k]} {{\tilde{ρ}}_{2 [k]} I_{n_{2}} + \frac{({\tilde{c}}_{12 [k]} h - 1) exp ({\tilde{c}}_{12 [k]} h) + 1}{{\tilde{c}}_{12 [k]}^{2}} {\tilde{a}}_{12 [k]} {\tilde{b}}_{12 [k]}^{T}} \frac{\partial}{\partial x_{1}} {\tilde{a}}_{12 [k]} \end{matrix}

(37)

\begin{matrix} = \frac{{\tilde{d}}_{12 [k]}}{| | {\tilde{x}}_{2 [k]} | |} {{\tilde{ρ}}_{2 [k]} I_{n_{2}} + \frac{({\tilde{c}}_{12 [k]} h - 1) exp ({\tilde{c}}_{12 [k]} h) + 1}{{\tilde{c}}_{12 [k]}^{2}} {\tilde{a}}_{12 [k]} {\tilde{b}}_{12 [k]}^{T}} \frac{\partial f_{2} (t, x_{1}, x_{2})}{\partial x_{1}} . \end{matrix}

(38)

Similarly, we obtain the

\frac{\partial x_{1}}{\partial x_{3}}

from Eq. (23).

\begin{matrix} \frac{\partial x_{1}}{\partial x_{3}} & = {\tilde{d}}_{11 [k]} {{\tilde{a}}_{11 [k]} \frac{\partial}{\partial x_{3}} {\tilde{ρ}}_{1 [k]} + {\tilde{ρ}}_{1 [k]} \frac{\partial}{\partial x_{3}} {\tilde{a}}_{11 [k]}} \end{matrix}

(39)

\begin{matrix} = {\tilde{d}}_{11 [k]} {{\tilde{ρ}}_{1 [k]} I_{n_{1}} + \frac{({\tilde{c}}_{11 [k]} h - 1) exp ({\tilde{c}}_{11 [k]} h) + 1}{{\tilde{c}}_{11 [k]}^{2}} {\tilde{a}}_{11 [k]} {\tilde{b}}_{11 [k]}^{T}} \frac{\partial}{\partial x_{3}} {\tilde{a}}_{11 [k]} \end{matrix}

(40)

\begin{matrix} = \frac{{\tilde{d}}_{11 [k]}}{| | {\tilde{x}}_{1 [k]} | |} {{\tilde{ρ}}_{1 [k]} I_{n_{1}} + \frac{({\tilde{c}}_{11 [k]} h - 1) exp ({\tilde{c}}_{11 [k]} h) + 1}{{\tilde{c}}_{11 [k]}^{2}} {\tilde{a}}_{11 [k]} {\tilde{b}}_{11 [k]}^{T}} \frac{\partial f_{1} (t, x_{1}, x_{2}, x_{3})}{\partial x_{3}} . \end{matrix}

(41)

Futhermore, we are able to get

\begin{matrix} J & = \frac{\partial f_{3}}{\partial x_{3}} = \frac{\partial f_{3}}{\partial x_{2}} \frac{\partial x_{2}}{\partial x_{1}} \frac{\partial x_{1}}{\partial x_{3}} \end{matrix}

(42)

\begin{matrix} = \frac{{\tilde{d}}_{12 [k]}}{| | {\tilde{x}}_{2 [k]} | |} \frac{{\tilde{d}}_{11 [k]}}{| | {\tilde{x}}_{1 [k]} | |} \frac{\partial f_{3}}{\partial x_{2}} E_{2} \frac{\partial f_{2}}{\partial x_{1}} E_{1} \frac{\partial f_{1}}{\partial x_{3}}, \end{matrix}

(43)

where

E_{i [k]} = {\tilde{ρ}}_{i [k]} I_{n_{i}} + \frac{({\tilde{c}}_{1 i [k]} h - 1) exp ({\tilde{c}}_{1 i [k]} h) + 1}{{\tilde{c}}_{1 i [k]}^{2}} {\tilde{a}}_{1 i [k]} {\tilde{b}}_{1 i [k]}^{T}

(

i = 1, 2

), and

F = U_{3} U_{2} U_{1}

(

U_{3} = \frac{\partial f_{3}}{\partial x_{2}}, U_{2} = \frac{\partial f_{2}}{\partial x_{1}}, U_{1} = \frac{\partial f_{1}}{\partial x_{3}}

) is nonsingular.

From Lemma 1, it is noted that

| E_{i [k]} | = {\tilde{ρ}}_{i [k]}^{n_{i} - 1} {{\tilde{ρ}}_{i [k]} + \frac{({\tilde{c}}_{1 i [k]} h - 1) exp ({\tilde{c}}_{1 i [k]} h) + 1}{{\tilde{c}}_{1 i [k]}^{2}} {\tilde{b}}_{1 i [k]}^{T} {\tilde{a}}_{1 i [k]}} = h {\tilde{ρ}}_{i [k]}^{n_{i} - 1} exp ({\tilde{c}}_{1 i [k]} h) \neq 0 .

For small

h < H

E_{i [k]} = h I_{n_{i}} + h^{2} {\tilde{a}}_{1 i [k]} {\tilde{b}}_{1 i [k]}^{T} + O (h^{3}) \approx h I_{n_{i}} + h^{2} {\tilde{a}}_{1 i [k]} {\tilde{b}}_{1 i [k]}^{T}

. Moreover, by using Lemma 3, if

h < min_{k} {\frac{1}{3 | | Q_{1 [k]} {| |}_{2} \cdot | | {\tilde{a}}_{11 [k]} {| |}_{2} \cdot | | {\tilde{b}}_{11 [k]} {| |}_{2}}, \frac{1}{3 | | Q_{2 [k]} {| |}_{2} \cdot | | {\tilde{a}}_{12 [k]} {| |}_{2} \cdot | | {\tilde{b}}_{12 [k]} {| |}_{2}}, \frac{1}{2 | | M_{k} {| |}_{2}}},

where

Q_{1 [k]} = U_{1} {(U_{3} E_{2} U_{2} U_{1})}^{- 1} U_{3} E_{2} U_{2}

Q_{2 [k]} = U_{2} U_{1} {(U_{3} U_{2} U_{1})}^{- 1} U_{3}

and

M_{k} = U_{1} {(U_{3} U_{2} U_{1})}^{- 1} U_{3} {\tilde{a}}_{12 [k]} {\tilde{b}}_{12 [k]}^{T} U_{2}

, then

U_{3} E_{2} U_{2} E_{1} U_{1}

is nonsingular. Therefore, satisfying for small h with

h < min_{k} {H, \frac{1}{3 | | Q_{1 [k]} {| |}_{2} \cdot | | {\tilde{a}}_{11 [k]} {| |}_{2} \cdot | | {\tilde{b}}_{11 [k]} {| |}_{2}}, \frac{1}{3 | | Q_{2 [k]} {| |}_{2} \cdot | | {\tilde{a}}_{12 [k]} {| |}_{2} \cdot | | {\tilde{b}}_{12 [k]} {| |}_{2}}, \frac{1}{2 | | M_{k} {| |}_{2}}},

we are able to obtain

J

is nonsingular. □

4. Numerical Examples

In this section, we do numerical simulation experiments to illustrate the performance of the MELGDAE. All codes are written under Windows 10 system and run on a personal computer with Intel(R) Core(TM) i5-3570 CPU @

3.40

GHz,

4.00

GB RAM and 64-bit operating system.

We consider a Hessenberg DAEs system with index 3 as follows, which can be found in [24]. Preprints 71017 i005

where

t \in [0, 1]

z_{1} (0) = z_{2} (0) = z_{3} (0) = z_{4} (0) = z_{5} (0) = 1

are the consistent initial values. And the exact solution is given by

z_{1} (t) = z_{3} (t) = e^{2 t}, z_{2} (t) = z_{4} (t) = e^{- t}, z_{5} (t) = e^{t} .

Let

x_{1} (t) = {(z_{1} (t), z_{2} (t))}^{T}, x_{2} (t) = {(z_{3} (t), z_{4} (t))}^{T}, x_{3} (t) = z_{5} (t)

, and

f_{1} = {(g_{1}, g_{2})}^{T}, f_{2} = {(g_{3}, g_{4})}^{T}, f_{3} = g_{5}

, then

F = \frac{\partial f_{3}}{\partial x_{2}} \frac{\partial f_{2}}{\partial x_{1}} \frac{\partial f_{1}}{\partial x_{3}} = 2 z_{2} (t) (2 z_{2} {(t)}^{2} z_{3} (t) + 1)

To obtain DAEs with index 2, we differentiate the constraint

g_{5}

to obtain the following equation.

0 = g_{6} = z_{1} (t) z_{4} (t) - z_{2} (t) z_{3} (t) .

(45)

Then, we are able to obtain index 2 DAEs below.

We compared the numerical results of the MELGDAE method with those of RADAU, a code based on implicit Runge-Kutta methods, and MEBDFI, a code based on the MEBDF method that has better theoretical properties than BDF methods, to demonstrate the capability of the MELGDAE method. The comparison was made using the Hessenberg DAEs with index 2 or 3 mentioned above.

Without loss of generality, we assume the absolute error (

ε_{1}

ε_{2}

ε_{3}

) is

10^{- 8}

, the relative error is

10^{- 5}

, the (initial) step size is

h = 10^{- 3}

θ_{1} = θ_{2} = θ_{3} = 0.5

. Based on Figure 1 and Figure 2, the MELGDAE method showed comparable accuracy with the RADAU method for the differential variables

z_{1} (t)

and

z_{3} (t)

for both index-2 and index-3 Hessenberg DAEs. In contrast, the MEBDFI method was less accurate for these variables. For the algebraic variable

z_{5} (t)

, both RADAU and MEBDFI methods provided more accurate results than the MELGDAE method. However, for the algebraic constraints

g_{5} (t)

and

g_{6} (t)

, the MELGDAE method exhibited significantly higher accuracy compared to the RADAU and MEBDFI methods.

Upon observing Figure 1 and Figure 2, it is clear that all three methods exhibit a reduction in the absolute errors of

z_{1} (t)

z_{3} (t)

, and

z_{5} (t)

as the index increases from 2 to 3. This reduction is not significant for the MELGDAE and RADAU methods. Therefore, we can conclude that the MELGDAE method has the potential to be extended for computing DAEs systems with an index higher than 3.

Furthermore, we conducted numerical experiments to verify the convergence accuracy of the MELGDAE method for index 3 DAEs. We assumed

h = {2^{- 10}, 2^{- 9}, 2^{- 8}, 2^{- 7}, 2^{- 6}, 2^{- 5}, 2^{- 4}}

and calculated their constants in

- l o g_{2} ({max}_{k} | z_{i [k]} - z_{i} (t_{k}) |) = ν (- l o g_{2} h) + μ

(

i = 1, 2, 3, 4, 5

) and

- l o g_{2} ({max}_{k} | g_{5} (t_{k}, z_{3 [k]}, z_{4 [k]}) |) = ν (- l o g_{2} h) + μ

, which represents

{max}_{k} | z_{i [k]} - z_{i} (t_{k}) | = O (h^{ν})

and

{max}_{k} | g_{5} (t_{k}, z_{3 [k]}, z_{4 [k]}) | = O (h^{ν})

, using the standard least-squares method. The results are presented in Figure 3, which indicate that the MELGDAE method exhibits second-order convergence for

z_{1} (t)

z_{2} (t)

z_{3} (t)

, and

z_{4} (t)

(all differential variables) and

g_{5}

(the algebraic constraints), and first-order convergence for

z_{5} (t)

(algebraic variable). Thus, this method may be extended to compute systems with an index of 4 or higher and is suitable for DAE systems that cannot be directly solved using other classical methods.

5. Conclusions

The Modified Extended Lie Group Differential Algebraic Equations (MELGDAE) method was developed in this paper to solve Hessenberg-DAEs with index 3. While the LGDAE method presented in [18] is only capable of handling index 2 systems, the MELGDAE method can solve higher index systems by taking into account the differential equation components with respect to the index in Hessenberg form. It is worth mentioning that the MELGDAE method demonstrates superior theoretical properties compared to the LGDAEs, as illustrated in Theorem 1.

We conducted a comparison of the MELGDAE method with the standard methods RADAU and MEBDFI on index 2 and 3 DAE systems. The MELGDAE method demonstrated promising results in terms of solution errors when compared to RADAU and MEBDFI, which are known to be powerful on different DAE systems. However, the MELGDAE method excelled in preserving the algebraic constraints. Moreover, all differential variables in index 3 Hessenberg-DAEs exhibited order 2 convergence with the MELGDAE method, whereas MEBDFI in [24] and RADAU in [7,25] demonstrated order reduction. Although the MELGDAE method has high stability and accuracy when solving strong nonlinear Hessenberg DAEs with index 3, there are still some theoretical gaps that need to be addressed in the future. Additionally, the MELGDAE method can be further extended to solve DAEs with an index of 4 or higher, which cannot be directly solved using standard methods.

Author Contributions

J.T. contributed to supervision, methodology, validation,funding acquisition and project administration. J.L. contributed to investigation, computations,funding acquisition and validation.

Funding

This work was partly supported by the NSF of China (No: 12201144), the Science and Technology Program of Guangzhou (No: 202002030138), the GuangDong Basic and Applied Basic Research Foundation (No: 2020A1515110554), the Science and Technology Foundation of Guizhou Province (No: QKHJC-ZK[2021]YB015), and the University-Industry Collaborative Education Program of the Ministry of Education (No: BINTECH-KJZX-20220831-61), China.

Acknowledgments

The authors wish to thank the referee for his or her very helpful comments and useful suggestions.

Conflicts of Interest

The authors declare no conflict of interest.

References

Schiehlen, W. Multibody system dynamics: roots and perspectives. Multibody System Dynamics 1997, 1, 149–188. [Google Scholar] [CrossRef]
Winkler, R. Stochastic differential algebraic equations of index 1 and applications in circuit simulation. Journal of computational and applied mathematics 2003, 157(2), 477–505. [Google Scholar] [CrossRef]
Coffey, T. S.; Kelley, C. T.; Keyes, D. E. Pseudotransient continuation and differential-algebraic equations. SIAM Journal on Scientific Computing 2003, 25(2), 553–569. [Google Scholar] [CrossRef]
Peng, H. J.; Li, F.; Liu, J. G.; Ju, Z. J. A symplectic instantaneous optimal control for robot trajectory tracking with differential-algebraic equation models. IEEE Transactions on Industrial Electronics 2019, 67(5), 3819–3829. [Google Scholar] [CrossRef]
Çelik, E.; Karaduman, E.; Bayram, M. Numerical solutions of chemical differential-algebraic equations. Applied Mathematics and Computation 2003, 139(2), 259–264. [Google Scholar] [CrossRef]
Ascher, U. M.; Petzold, L. R. Projected implicit Runge–Kutta methods for differential-algebraic equations. SIAM Journal on Numerical Analysis 1991, 28(4), 1097–1120. [Google Scholar] [CrossRef]
Hairer, E.; Wanner, G. Stiff differential equations solved by radau methods. Journal of Computational and Applied Mathematics 1999, 111, 93–111. [Google Scholar] [CrossRef]
Skvortsov, L. M. Diagonally implicit Runge—Kutta methods for differential algebraic equations of indices two and three. Computational Mathematics and Mathematical Physics 2010, 50, 993–1005. [Google Scholar] [CrossRef]
Cash, J.; Considine, S. An mebdf code for stiff initial value problems. ACM Trans. Math. Softw. 1992, 142–158. [Google Scholar] [CrossRef]
Cash, J. R. Modified extended backward differentiation formulae for the numerical solution of stiff initial value problems in ODEs and DAEs. Journal of Computational and Applied Mathematics 2000, 125(1), 117–130. [Google Scholar] [CrossRef]
Karta, M.; Celik, E. On the numerical solution of differential-algebraic equations with Hessenberg Index-3. Discrete Dynamics in Nature and Society 2012, 2012, 12. [Google Scholar] [CrossRef]
Benhammouda, B. A novel technique to solve nonlinear higher-index Hessenberg differential-algebraic equations by Adomian decomposition method. SpringerPlus 2016, 5(1), 1–14. [Google Scholar] [CrossRef]
Celledoni, E.; Kometa, B. K. Semi-Lagrangian multistep exponential integrators for index 2 differential–algebraic systems. Journal of Computational Physics 2011, 230(9), 3413–3429. [Google Scholar] [CrossRef]
Iserles, A.; Zanna, A. Preserving algebraic invariants with Runge–Kutta methods. Journal of Computational and Applied Mathematics 2000, 125(1), 69–81. [Google Scholar] [CrossRef]
Wieloch, V.; Arnold, M. BDF integrators for constrained mechanical systems on Lie groups. Journal of Computational and Applied Mathematics 2021, 387, 112517. [Google Scholar] [CrossRef]
Liu, C. S. Preserving constraints of differential equations by numerical methods based on integrating factors. Computer Modeling in Engineering and Sciences 2006, 12(2), 83–107. [Google Scholar]
Liu, C. S. Finding unknown heat source in a nonlinear Cauchy problem by the Lie-group differential algebraic equations method. Engineering Analysis with Boundary Elements 2015, 50, 148–156. [Google Scholar] [CrossRef]
Liu, C. S.; Chen, L.; Liu, L. W. Solving mechanical systems with nonholonomic constraints by a Lie-group differential algebraic equations method. Journal of Engineering Mechanics 2017, 143(9), 04017097. [Google Scholar] [CrossRef]
Liu, C. S. Lie-group differential algebraic equations method to recover heat source in a Cauchy problem with analytic continuation data. International Journal of Heat and Mass Transfer 2014, 78, 538–547. [Google Scholar] [CrossRef]
Liu, C. S.; Kuo, C. L.; Chang, J. R. Recovering a heat source and initial value by a Lie-group differential algebraic equations method. Numerical Heat Transfer, Part B: Fundamentals 2015, 67(3), 231–254. [Google Scholar] [CrossRef]
Liu, C. S. A new sliding control strategy for nonlinear system solved by the Lie-group differential algebraic equation method. Communications in Nonlinear Science and Numerical Simulation 2014, 19(6), 2012–2038. [Google Scholar] [CrossRef]
Liu, C. S.; Chang, C. W. A real-time Lie-group differential algebraic equations method to solve the inverse nonlinear vibration problems. Inverse Problems in Science and Engineering 2016, 24(9), 1569–1586. [Google Scholar] [CrossRef]
Horn, R.A.; Johnson, C.R. Matrix Analysis, 2nd ed.; Cambridge University Press: New York, USA, 2012; pp. 66–67. [Google Scholar]
Jay, L. Convergence of Runge-Kutta methods for differential-algebraic systems of index 3. Applied Numerical Mathematics 1995, 17(2), 97–118. [Google Scholar] [CrossRef]
Hairer, E.; Lubich, C.; Roche, M. The numerical solution of differential-algebraic systems by Runge-Kutta methods; Springer-Verlag: Berlin, Germany, 1989; pp. 19–20. [Google Scholar]

Figure 1. The abosulte errors of three methods for index -2 DAEs. (a1)-(a4) are the absolute errors of

z_{1} (t)

z_{3} (t)

z_{5} (t)

and

g_{6} (t)

of RADAU respectively; (b1)-(b4) are the absolute errors of MEBDFI; (c1)-(c4) are the absolute errors of MELGDAE.

Figure 1. The abosulte errors of three methods for index -2 DAEs. (a1)-(a4) are the absolute errors of

z_{1} (t)

z_{3} (t)

z_{5} (t)

and

g_{6} (t)

of RADAU respectively; (b1)-(b4) are the absolute errors of MEBDFI; (c1)-(c4) are the absolute errors of MELGDAE.

Figure 2. The abosulte errors of three methods for Index 3 DAEs. (a1)-(a4) are the absolute errors of

z_{1} (t)

z_{3} (t)

z_{5} (t)

and

g_{5} (t)

of RADAU respectively; (b1)-(b4) are the absolute errors of MEBDFI; (c1)-(c4) are the absolute errors of MELGDAE.

Figure 2. The abosulte errors of three methods for Index 3 DAEs. (a1)-(a4) are the absolute errors of

z_{1} (t)

z_{3} (t)

z_{5} (t)

and

g_{5} (t)

of RADAU respectively; (b1)-(b4) are the absolute errors of MEBDFI; (c1)-(c4) are the absolute errors of MELGDAE.

Figure 3. The convergence order

O (h^{ν})

of MELGDAE for index 3 DAEs.

Figure 3. The convergence order

O (h^{ν})

of MELGDAE for index 3 DAEs.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.

MDPI Initiatives

Important Links

Choose an area of interest and we will send you notifications of new preprints at your preferred frequency.

Disclaimer

Modified Extended Lie Group Method for Hessenberg Differential Algebraic Equations with Index 3

Abstract

1. Introduction

2. Preliminaries

2.1. Hessenberg DAEs

2.2. Lie group structure of ODEs

2.3. An implicit G L ( n , R ) Lie-group method

3. MELGDAE method for index-3 Hessenberg-DAEs

4. Numerical Examples

5. Conclusions

Author Contributions

Funding

Acknowledgments

Conflicts of Interest

References

MDPI Initiatives

Important Links

Subscribe

2.3. An implicit $G L (n, R)$ Lie-group method