Next Article in Journal
Risk Level Evaluation on Construction Project Lifecycle Using Fuzzy Comprehensive Evaluation and TOPSIS
Next Article in Special Issue
Symplectic Effective Order Numerical Methods for Separable Hamiltonian Systems
Previous Article in Journal
On the Evolutionary Form of the Constraints in Electrodynamics
Previous Article in Special Issue
Improvement of Risk Assessment Using Numerical Analysis for an Offshore Plant Dipole Antenna

Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

# A Complex Lie-Symmetry Approach to Calculate First Integrals and Their Numerical Preservation

by
1,
Yousaf Habib
1,2,* and
3
1
School of Natural Sciences, National University of Sciences and Technology, Sector H-12, Islamabad 44000, Pakistan
2
Department of Mathematics, COMSATS University Islamabad, Lahore Campus, Lahore 54000, Pakistan
3
Department of BS and H, College of E and ME, National University of Sciences and Technology, Peshawar Road, Rawalpindi 46000, Pakistan
*
Author to whom correspondence should be addressed.
Symmetry 2019, 11(1), 11; https://doi.org/10.3390/sym11010011
Submission received: 26 September 2018 / Revised: 24 October 2018 / Accepted: 31 October 2018 / Published: 24 December 2018
(This article belongs to the Special Issue Numerical Analysis or Numerical Method in Symmetry)

## Abstract

:
We calculated Noether-like operators and first integrals of a scalar second-order ordinary differential equation using the complex Lie-symmetry method. We numerically integrated the equations using a symplectic Runge–Kutta method. It was seen that these structure-preserving numerical methods provide qualitatively correct numerical results, and good preservation of first integrals is obtained.
MSC:
34C14; 37K05; 70G65; 65L05; 65L06

## 1. Introduction

Marius Sophus Lie proposed a symmetry-based method for the analytical solution of differential equations using groups of continuous transformations known as Lie groups [1,2,3,4]. Amalie Emmy Noether later presented her remarkable theorem that relates variational symmetries with conservation laws or first integrals in Reference [5]. In the literature, different methods are available to calculate first integrals of ordinary differential equations (ODEs), including the direct method, the characteristic or multiplier method, the Noether approach, and the partial Noether approach [6,7,8,9]. In this paper, we used the classical Noether approach to calculate the first integrals of a harmonic oscillator. We then applied the complex symmetry method in the restricted domain to find the first integrals of a system of harmonic oscillators by considering the Lagrangian in the complex variable domain [10,11,12].
Concerning the numerical solutions of ODEs with quadratic first integrals, it is well known that symplectic numerical methods are a suitable candidate [13]. These methods are a subclass of geometric integrators that preserve the geometric properties of the exact flow of ODEs. One class of symplectic methods with optimal order are the Gauss–Legendre Runge–Kutta methods. They are one-step numerical methods for ODEs and preserve all linear and quadratic first integrals of a dynamic system [14]. If we intend to preserve cubic or higher-order first integrals, we do not have a general numerical scheme for such a purpose, but we can design a numerical method that has this as its specific goal, for example, with the splitting and discrete-gradient methods [14]. In this paper, we present a way of constructing symplectic Runge–Kutta methods. We then take fourth-order Gauss–Legendre Runge–Kutta methods for the numerical integration of ODEs and report good preservation of first integrals by the numerical solution.

## 2. Symmetries and First Integrals

Consider a second-order ordinary differential equation,
$d 2 y d t = f ( t , y ′ , y ) ,$
which admits a Lagrangian L satisfying the Euler–Lagrange equation,
$d d t ( ∂ L ∂ y ′ ) − ∂ L ∂ y = 0 .$
To explain the invariance criteria for variational problems under a group of transformation, we consider the operator
$X = ξ ( t , y ) ∂ ∂ t + η ( t , y ) ∂ ∂ y ,$
where X is the Noether symmetry generator for the Lagrangian L with gauge function $B ( t , y )$, provided the following condition holds,
$X ( 1 ) ( L ) + D t ( ξ ) L − D t ( B ) = 0 ,$
where $X ( 1 )$ is a first-order prolongation of X and D represents total derivative,
$D t = ∂ ∂ t + y ′ ∂ ∂ y .$
According to the Noether theorem, for each Noether symmetry of a Euler–Lagrange equation, there corresponds a function I
$I = ξ L + ( η − ξ y ′ ) ∂ L ∂ y ′ − B ( t , y ) ,$
called the first integral or conserved quantity of Equation (1) with respect to symmetry generator X.

#### Complex Symmetry Analysis

We first discuss some important results related to complex Noether symmetries, complex Lagrangian, and the Noether theorem in the restricted complex domain. We use them to determine first integrals of second-order restricted complex ODEs [15]. We then present expressions for Euler–Lagrange-like equations, conditions for Noether-like operators, and expressions for first integrals corresponding to these operators. For more details, see Reference [10] and references therein.
Consider a system of two second-order ordinary differential equations of the form
$d 2 f d t = w 1 ( t , g , f , g ′ , f ′ ) , d 2 g d t = w 2 ( t , g , f , g ′ , f ′ ) .$
Suppose we have a transformation $y ( t ) = f + i g$ and $w = w 1 + i w 2$, which converts System (7) to a second-order restricted complex ODE,
$y ″ = w ( t , y , y ′ ) .$
Assume that Equation (8) admits a complex Lagrangian $L ( t , f , g , f ′ , g ′ )$, i.e., $L = L 1 + i L 2$. Therefore, we have two Lagrangians, $L 1$ and $L 2$, for System (7) that satisfy Euler–Lagrange-like equations:
$∂ L 1 ∂ f + ∂ L 2 ∂ g − d d t ( ∂ L 1 ∂ f ′ + ∂ L 1 ∂ g ′ ) = 0 , ∂ L 2 ∂ f − ∂ L 1 ∂ g − d d t ( ∂ L 2 ∂ f ′ − ∂ L 1 ∂ g ′ ) = 0 .$
The operators
$X 1 = ς 1 ∂ ∂ t + χ 1 ∂ ∂ f + χ 2 ∂ ∂ g , X 2 = ς 2 ∂ ∂ t + χ 2 ∂ ∂ f − χ 1 ∂ ∂ g .$
are called Noether-like operators for Lagrangians $L 1$ and $L 2$ such that:
$− X 2 ( 1 ) ( L 2 ) + X 1 ( 1 ) ( L 1 ) + ( D t ς 1 ) L 1 − ( D t ς 2 ) L 2 = D t A 1 , X 2 ( 1 ) ( L 1 ) + X 1 ( 1 ) ( L 2 ) + ( D t ς 1 ) L 2 + ( D t ς 2 ) L 1 = D t A 2 ,$
where $A 1$ and $A 2$ are suitable gauge functions. The two first integrals corresponding to Noether-like operators $X 1$ and $X 2$ can be found as:
$I 1 = − A 1 + ς 1 L 1 + ∂ f ′ L 1 ( χ 1 − ς 2 L 2 − ς 1 f ′ − ς 2 g ′ ) − ∂ f ′ L 2 ( χ 2 − ς 2 f ′ − ς 1 g ′ ) , I 2 = − A 2 + ς 1 L 2 + ∂ f ′ L 2 ( χ 1 + ς 2 L 1 − ς 1 f ′ − ς 2 g ′ ) + ∂ f ′ L 1 ( χ 2 − ς 2 f ′ − ς 1 g ′ ) .$

## 3. Runge–Kutta Methods

Runge–Kutta methods [16] are one-step numerical methods for the approximate solution of IVPs:
$y ′ ( t ) = f ( y ( t ) ) , y ( t 0 ) = y 0 , y ( t ) ∈ R n .$
These methods provide approximation $y n = y ( t n )$ of the exact solution $y ( t )$ at time $t n = n h$, where $n = 0 , 1 , …$ and h corresponds to the stepsize. The generalized form of an s-stage Runge–Kutta method is
$Y k = y n − 1 + ∑ i = 1 s a k i h f ( Y i ) , k = 1 , … , s , y n = y n − 1 + ∑ l = 1 s b l h f ( Y l ) ,$
with $b i$ representing the weights and $c i$, the nodes at which stages $Y k$ are evaluated. A Runge–Kutta method can be represented by a Butcher tableau:
$c 1 a 11 ⋯ a 1 n ⋮ ⋮ ⋱ ⋮ c n a n 1 ⋯ a n n b 1 ⋯ b n .$
For explicit Runge–Kutta methods, we have $a k i = 0$ for $k ≤ i ,$; otherwise, they are implicit.

#### 3.1. Symplectic Runge–Kutta Methods

If Equation (13) has a quadratic first integral
$I ( y ) = 〈 y , S y 〉 = y T S y ,$
where S is a symmetric square matrix, then we have
$〈 y , f ( y ) 〉 = y T S f ( y ) = 0 .$
We want to determine numerical solutions $y n$ such that first integral $I ( y )$ is preserved numerically, i.e.,
$〈 y n , S y n 〉 = 〈 y n − 1 , S y n − 1 〉 n = 0 , 1 , … .$
It has been shown in References [17,18,19] that only symplectic Runge–Kutta methods preserve quadratic first integrals while numerically integrating System (13). Moreover, in this paper we are only considering implicit Runge–Kutta methods to check the numerical preservation of first integrals because explicit methods cannot be symplectic [20]. A Runge–Kutta method is symplectic if its coefficients satisfy the following condition [18,19,21]:
$b i a i j + b j a j i − b i b j = 0 ∀ j , i = 1 , … , s ,$
which can be derived as follows.
Firstly, apply the Runge–Kutta method (14) to solve the IVP (13). The stage values are
$Y i = y n − 1 + ∑ j h a i j f ( Y j ) .$
Since
$〈 Y i , S f ( Y i ) 〉 = 0 , ⇒ 〈 y n − 1 , S f ( Y i ) 〉 + ∑ j h a i j 〈 f ( Y j ) , S f ( Y i ) 〉 = 0 .$
Moreover, for the output values, we have
$y n = y n − 1 + ∑ i = 1 s b i h f ( Y i ) .$
Thus,
$〈 y n , S y n 〉 = 〈 y n − 1 , S y n − 1 〉 + h ∑ i b i 〈 y n − 1 , S f ( Y i ) 〉 + h ∑ j b j 〈 f ( Y j ) , S y n − 1 〉 + h 2 ∑ i , j b j b i 〈 f ( Y i ) , S f ( Y j ) 〉 .$
Evidently from Systems (16) and (17), we have
$〈 y n , S y n 〉 = 〈 y n − 1 , S y n − 1 〉 ,$
provided that
$b j a j i + b i a i j − b i b j = 0 .$

#### 3.2. Construction of Symplectic RK Methods

Although there exist several techniques to construct symplectic RK methods in the literature [14,22], here we constructed symplectic Runge–Kutta methods with the help of a Vandermonde transformation. This was first discussed in reference [23].
A Vandermonde matrix is given as
$V = 1 c 1 ⋯ c 1 n − 1 ⋮ ⋮ ⋱ ⋮ 1 c n ⋯ c n n − 1 = c i j − 1 .$
Pre- and postmultiply, Vandermonde matrix V with symplectic condition (15) as
$c i k − 1 ( b j a j i + b i a i j − b j b i ) c j l − 1 = 0 , ∀ l , k , j , i = 1 , 2 , … , s .$
To construct methods with two stages ($s = 2$), we consider
For $l , k = 1$,
$∑ i , j ( b j a j i + b i a i j − b j b i ) = 0 .$
For $l = 1$ and $k = 2 ,$
$∑ i , j ( b j c j a j i + b i a i j c j − b j c j b i ) = 0 .$
For $l = 2$ and $k = 1 ,$
$∑ i , j ( b i a i j c j + b j c j a j i − b i b j c j ) = 0 .$
For $l , k = 2$,
$∑ i , j ( b i c i a i j c j + b j c j a j i c i − b i c i b j c j ) = 0 .$
The following order two conditions must be satisfied.
$∑ j b j = 1 , ∑ j b j c j = 1 2 .$
Using Equation (24) in Equations (20)–(23), we have
$∑ i b i c i = 1 2 , ∑ i , j = 1 2 ( b i a i j c j + b j c j a j i ) = 1 2 , ∑ i , j = 1 2 ( b i c i a i j + b j a j i c i ) = 1 2 , ∑ i , j = 1 2 b i c i a i j c j = 1 8 .$
If we take $b i ( c i − c 1 ) = b i c i − b i c 1 )$ and take summation of i from 1 to 2, we get
$b 2 c 2 − b 2 c 1 = 1 2 − c 1 ,$
$b 2 = c 1 − 1 2 c 1 − c 2 .$
Similarly,
$b 1 = c 2 − 1 2 c 2 − c 1 .$
If we take the relation
$b i ( c j − c 1 ) a i j ( c i − c 1 ) = b i c i a i j c j − b i c i a i j c 1 − b i a i j c j c 1 + b i a i j c 1 c 1 .$
Thus, we get
$a 22 = 1 8 − c 1 6 − c 1 3 + c 1 c 1 2 b 2 ( c 2 − c 1 ) 2 .$
Similarly, we get
$a 11 = 1 8 − c 2 6 − c 2 3 + c 2 c 2 2 b 1 ( c 1 − c 2 ) 2 ,$
$a 21 = 1 8 − c 2 3 − c 1 6 + c 1 c 2 2 b 2 ( c 2 − c 1 ) ( c 1 − c 2 ) ,$
$a 12 = 1 8 − c 1 3 − c 2 6 + c 1 c 2 2 b 1 ( c 2 − c 1 ) ( c 1 − c 2 ) .$
Let us consider the shifted Legendre polynomials $P t *$ on the interval $[ 0 , 1 ]$,
$P t * ( y ) = ∑ n = 0 t t ! 2 t t n t + n n ( − 1 ) t − n y n .$
For Gauss methods, we choose abscissa $c i$ as zeros of $P t *$ which have an order $2 t$. For Radau methods, we choose either $c 1 = 0$ or $c t = 1$, or both of them and then take for Radau I methods, the abscissa as the zeros of the polynomial $P t − 1 * ( y ) + P t * ( y )$ of order $2 t − 1$. Similarly, for Radau II methods, we take the abscissa as the zeros of the polynomial $P t * ( y ) − P t − 1 * ( y )$ of order $2 t − 1$. Moreover, for Lobatto III methods, we take the abscissa as the zeros of the polynomial $P t * ( y ) − P t − 2 * ( y )$ of order $2 t − 2$. Thus, we have the following symplectic methods:
Gauss, s = 2:
$1 2 − 3 6 1 4 1 4 − 3 6 1 2 + 3 6 1 4 + 3 6 1 4 1 2 1 2$
$0 1 8 − 1 8 2 3 7 24 3 8 1 4 3 4$
$1 3 3 8 − 1 24 1 7 8 1 8 3 4 1 4$
Similarly, we can construct methods with more stages and a higher order.

## 4. Construction of First Integrals and Their Numerical Preservation

We construct the first integrals of a system of harmonic oscillators (both coupled and uncoupled) determined by the second-order ODE:
$y ″ = − k 2 y .$
We take different values of k and y, as follows:
Case I: ($k 2 = 1$ and $y$ is real)
When $k 2 = 1$ and $y ( t )$ is real-valued, (25) becomes a one-dimensional harmonic oscillator equation:
$y ″ = − y ,$
that possesses the standard Lagrangian
$L = y ′ 2 2 − y 2 2 .$
Taking the Lagrangian and inserting it in System (4) yields the following determining system of equations:
$− η y + η t y ′ + ( η y − 1 2 ξ t ) y ′ 2 − 1 2 ξ y y ′ 3 − 1 2 ξ t y 2 − 1 2 ξ y y 2 y ′ − B t − y ′ B y = 0 .$
Comparing different powers of $y ′$, we have a system of four partial differential equations whose solution gives rise to:
$ξ ( t , y ) = c 1 + c 2 sin ( 2 t ) + c 3 cos ( 2 t ) , η ( t , y ) = sin t c 4 + ( cos ( 2 t ) y c 2 − sin ( 2 t ) y c 3 ) + cos t c 5 , B ( t , y ) = − ( c 2 sin ( 2 t ) + c 3 cos ( 2 t ) ) y 2 + ( c 4 cos t − c 5 sin t ) y .$
We thus obtain the following 5-Noether symmetry generators:
$X 1 = ∂ ∂ t , X 2 = sin ( 2 t ) ∂ ∂ t + y cos ( 2 t ) ∂ ∂ y , X 3 = cos ( 2 t ) ∂ ∂ t − y sin ( 2 t ) ∂ ∂ y , X 4 = cos ( t ) ∂ ∂ y , X 5 = sin ( t ) ∂ ∂ y .$
Using Symmetries (30) and Lagrangian (27) in Noether’s Theorem (6), we obtain the following first integrals:
$I 1 = y ′ 2 2 + y 2 2 , I 2 = y sin t + y ′ cos t , I 3 = − y cos t + y ′ sin t , I 4 = − 1 2 y ′ 2 cos 2 t − y y ′ sin 2 t + 1 2 y 2 cos 2 t , I 5 = − 1 2 y ′ 2 sin 2 t + y y ′ cos 2 t + 1 2 y 2 sin 2 t .$
Among these five first integrals, only two are independent [8]. We numerically integrate system (26) using a fourth-order Gauss $s = 2$ symplectic Runge–Kutta method that we refer to from now on as Gauss2. We compare the results of the Gauss2 method with the famous symplectic Euler method [14], given as:
$U n + 1 = U n + h f ( V n ) , V n + 1 = V n − h g ( U n + 1 ) ,$
for numerically integrating $U ′ = f ( V )$ and $V ′ = g ( U )$. We take stepsize $h = 0.01$, and n = 10,000 number of steps. By employing symplectic integrators, we expect the first integrals of the system to be preserved by the numerical schemes, and this is what we have achieved. We look at the deviation of numerically evaluated first integral $I ( y n )$ from the actual value of first integral $I ( y 0 )$. We calculate error by taking the difference of the first integral evaluated at initial value $I ( y 0 )$ with the value of the first integral evaluated at all subsequent numerically approximated values $I ( y n )$ given by the formula Error $= | I ( y n ) − I ( y 0 ) |$. Figure 1 and Figure 2 represent the absolute error in integral $I 2$ using the Gauss2 and symplectic Euler method, respectively. Similarly, Figure 3 and Figure 4 represent the absolute error in integral $I 3$ using the Gauss2 and symplectic Euler method, respectively. It is clear from the figures that the error is very small and bounded, depicting qualitatively correct numerical results. It is worth noting that the error of the Gauss2 method is much less compared to the error of the symplectic Euler method. The reason is that the Gauss2 method is fourth-order and more accurate compared to symplectic Euler method, which has order 1. Similar error behavior is obtained for other first integrals.
Case II: ($k 2 = 1$ and $y$ is complex)
When $k 2 = 1$ and $y ( t )$ is a complex function $y = f + i g$ for f and g being real functions of t, we get:
$f ″ = − f , g ″ = − g ,$
$L 1 = 1 2 ( − g ′ g ′ + f ′ f ′ − f f + g g ) , L 2 = g ′ f ′ − f g .$
Using Lagrangians (33) in (11), we obtain 9-Noether-like operators
$X 1 = ∂ ∂ t , X 2 = sin t ∂ ∂ f , X 3 = sin t ∂ ∂ g , X 4 = cos t ∂ ∂ f , X 5 = cos t ∂ ∂ g , X 6 = sin 2 t ∂ ∂ t + f cos 2 t ∂ ∂ f + g cos 2 t ∂ ∂ g , X 7 = g cos 2 t ∂ ∂ f − f cos 2 t ∂ ∂ g , X 8 = cos 2 t ∂ ∂ t − f sin 2 t ∂ ∂ f − g sin 2 t ∂ ∂ g , X 9 = − g sin 2 t ∂ ∂ f + f sin 2 t ∂ ∂ g .$
Invoking Equation (12), we obtain the following invariants:
$I 1 , 1 = ( f ′ f ′ − g ′ g ′ − f f + g g ) sin 2 t − 2 ( f f ′ − g g ′ ) cos 2 t , I 1 , 2 = 2 ( f ′ g ′ − f g ) sin 2 t − 2 ( f g ′ + f ′ g ) cos 2 t , I 2 , 1 = ( f ′ f ′ − g ′ g ′ − f f + g g ) cos 2 t + 2 ( f ′ f − g ′ g ) sin 2 t , I 2 , 2 = 2 ( f ′ g ′ − f g ) cos 2 t + 2 ( f g ′ + f ′ g ) sin 2 t , I 3 , 1 = − 2 f ′ cos t − 2 f sin t , I 3 , 2 = − 2 g ′ cos t − 2 g sin t , I 4 , 1 = − 2 f ′ sin t + 2 f cos t , I 4 , 2 = − 2 g ′ sin t + 2 g cos t , I 5 , 1 = f ′ f ′ − g ′ g ′ − f 2 + g 2 , I 5 , 2 = 2 g ′ f ′ + 2 g f .$
associated with Noether-like operators (34). System of Equation (32) is integrated using the Gauss2 method with stepsize $h = 0.01$ and n = 10,000 number of steps. The absolute error in first integrals $I 2 , 1$, $I 2 , 2$, $I 4 , 1$, and $I 4 , 2$ is plotted in Figure 5, Figure 6, Figure 7 and Figure 8, respectively. Similar error behavior is obtained for $I 1 , 1$, $I 1 , 2$, $I 3 , 1$, $I 3 , 2$, $I 5 , 1$, and $I 5 , 2$. We observe that the error does not grow out of bounds, which shows that the numerical method can mimic the true qualitative feature of the dynamical system.
Case III: ($k$ and $y$ are complex)
When k and $y ( t )$ are both complex, i.e., $k = α 1 + i α 2$ and $y = f + i g$ for f, g, $α 1$, and $α 2$ being real, the following coupled system of harmonic oscillators is obtained:
$f ″ = − ( α 1 2 − α 2 2 ) f + 2 α 1 α 2 g , g ″ = − ( α 1 2 − α 2 2 ) g − 2 α 1 α 2 f ,$
which admits a pair of Lagrangians [11]:
$L 1 = 1 2 f ′ 2 − 1 2 g ′ 2 − 1 2 ( α 1 2 − α 2 2 ) ( f 2 − g 2 ) + 2 α 1 α 2 f g , L 2 = f ′ g ′ − α 1 α 2 ( f 2 − g 2 ) − ( α 1 2 − α 2 2 ) f g .$
System (36) admits the following 9 Noether-like operators and first integrals:
$X 1 = ∂ ∂ t , X 2 = sin ( α 1 t ) cosh ( α 2 t ) ∂ ∂ f + cos ( α 1 t ) sinh ( α 2 t ) ∂ ∂ g , X 3 = cos ( α 1 t ) sinh ( α 2 t ) ∂ ∂ f − sin ( α 1 t ) cosh ( α 2 t ) ∂ ∂ g , X 4 = cos ( α 1 t ) cosh ( α 2 t ) ∂ ∂ f − sin ( α 1 t ) sinh ( α 2 t ) ∂ ∂ g , X 5 = − sin ( α 1 t ) sinh ( α 2 t ) ∂ ∂ f − cos ( α 1 t ) cosh ( α 2 t ) ∂ ∂ g , X 6 = sin ( 2 α 1 t ) cosh ( 2 α 2 t ) ∂ ∂ t + { ( α 1 f − α 2 g ) cos ( 2 α 1 t ) cosh ( 2 α 2 t ) + { ( α 1 g + α 2 f ) cos ( 2 α 1 t ) cosh ( 2 α 2 t ) − ( α 1 f − α 2 g ) sin ( 2 α 1 t ) sinh ( 2 α 2 t ) } ∂ ∂ g + ( α 1 g + α 2 f ) sin ( 2 α 1 t ) sinh ( 2 α 2 t ) } ∂ ∂ f , X 7 = cos ( 2 α 1 t ) sinh ( 2 α 2 t ) ∂ ∂ t + { ( α 1 g + α 2 f ) cos ( 2 α 1 t ) cosh ( 2 α 2 t ) − { ( α 1 f − α 2 g ) cos ( 2 α 1 t ) cosh ( 2 α 2 t ) + ( α 1 g + α 2 f ) sin ( 2 α 1 t ) sinh ( 2 α 2 t ) } ∂ ∂ g − ( α 1 f − α 2 g ) sin ( 2 α 1 t ) sinh ( 2 α 2 t ) } ∂ ∂ f , X 8 = cos ( 2 α 1 t ) cosh ( 2 α 2 t ) ∂ ∂ t + { ( α 1 f − α 2 g ) sin ( 2 α 1 t ) cosh ( 2 α 2 t ) + { ( α 1 f − α 2 g ) cos ( 2 α 1 t ) sinh ( 2 α 2 t ) + ( α 1 g + α 2 f ) sin ( 2 α 1 t ) cosh ( 2 α 2 t ) } ∂ ∂ g − ( α 1 g + α 2 f ) cos ( 2 α 1 t ) sinh ( 2 α 2 t ) } ∂ ∂ f , X 9 = − sin ( 2 α 1 t ) sinh ( 2 α 2 t ) ∂ ∂ t + { ( α 1 f − α 2 g ) cos ( 2 α 1 t ) sinh ( 2 α 2 t ) − { ( α 1 f − α 2 g ) sin ( 2 α 1 t ) cosh ( 2 α 2 t ) − ( α 1 g + α 2 f ) cos ( 2 α 1 t ) sinh ( 2 α 2 t ) } ∂ ∂ g + ( α 1 g + α 2 f ) sin ( 2 α 1 t ) cosh ( 2 α 2 t ) } ∂ ∂ f .$
$I 1 , 1 = ( α 1 2 − α 2 2 ) ( f 2 − g 2 ) + f ′ 2 − 4 α 1 α 2 f g − g ′ 2 , I 1 , 2 = 2 ( α 1 2 − α 2 2 ) f g + 2 α 1 α 2 ( f 2 − g 2 ) + 2 f ′ g ′ , I 2 , 1 = f ′ sin ( α 1 t ) cosh ( α 2 t ) − g ′ cos ( α 1 t ) sinh ( α 2 t ) − ( α 1 f − α 2 g ) cos ( α 1 t ) cosh ( α 2 t ) − ( α 1 g + α 2 f ) sin ( α 1 t ) sinh ( α 2 t ) , I 2 , 2 = g ′ sin ( α 1 t ) cosh ( α 2 t ) + f ′ cos ( α 1 t ) sinh ( α 2 t ) − ( α 1 g + α 2 f ) cos ( α 1 t ) cosh ( α 2 t ) + ( α 1 f − α 2 g ) sin ( α 1 t ) sinh ( α 2 t ) , I 3 , 1 = f ′ cos ( α 1 t ) cosh ( α 2 t ) + g ′ sin ( α 1 t ) sinh ( α 2 t ) + ( α 1 f − α 2 g ) sin ( α 1 t ) cosh ( α 2 t ) − ( α 1 g + α 2 f ) cos ( α 1 t ) sinh ( α 2 t ) , I 3 , 2 = g ′ cos ( α 1 t ) cosh ( α 2 t ) − f ′ sin ( α 1 t ) sinh ( α 2 t ) + ( α 1 g + α 2 f ) sin ( α 1 t ) cosh ( α 2 t ) + ( α 1 f − α 2 g ) cos ( α 1 t ) sinh ( α 2 t ) , I 4 , 1 = 1 2 [ { ( α 1 2 − α 2 2 ) ( − g 2 + f 2 ) − 4 α 1 α 2 g f − ( − g ′ 2 + f ′ 2 ) } sin ( 2 α 1 t ) cosh ( 2 α 2 t ) − { 2 α 1 α 2 f 2 − 2 α 1 α 2 g 2 + 2 ( α 1 2 − α 2 2 ) g f − 2 g ′ f ′ } cos ( 2 α 1 t ) sinh ( 2 α 2 t ) ] + { α 1 ( f ′ f − g ′ g ) − α 2 ( g ′ f + f ′ g ) } cos ( 2 α 1 t ) cosh ( 2 α 2 t ) + { α 1 g ′ f + α 1 f ′ g + α 2 f ′ f − α 2 g ′ g } sin ( 2 α 1 t ) sinh ( 2 α 2 t ) I 4 , 2 = 1 2 [ { ( α 1 2 − α 2 2 ) ( − g 2 + f 2 ) − 4 α 1 α 2 g f + ( g ′ 2 − f ′ 2 ) } cos ( 2 α 1 t ) sinh ( 2 α 2 t ) + { 2 α 1 α 2 ( f 2 − g 2 ) + 2 f g ( α 1 2 − α 2 2 ) − 2 f ′ g ′ } sin ( 2 α 1 t ) cosh ( 2 α 2 t ) ] + [ { α 1 f ′ g + α 1 f g ′ + α 2 ( f ′ f − g g ′ ) } cos ( 2 α 1 t ) cosh ( 2 α 2 t ) ] − [ { α 1 f f ′ − α 1 g ′ g ) − α 2 g ′ f − α 2 f ′ g ) } sin ( 2 α 1 t ) sinh ( 2 α 2 t ) ] I 5 , 1 = 1 2 [ { ( α 1 2 − α 2 2 ) ( − g 2 + f 2 ) − 4 α 1 α 2 f g − ( − g ′ 2 + f ′ 2 ) } cos ( 2 α 1 t ) cosh ( 2 α 2 t ) + { 2 α 1 α 2 f 2 − { 2 α 1 α 2 g 2 + 2 ( α 1 2 − α 2 2 ) g f − 2 g ′ f ′ } sin ( 2 α 1 t ) sinh ( 2 α 2 t ) ] + { α 1 f g ′ + α 1 g f ′ + α 2 ( f f ′ − g g ′ ) } cos ( 2 α 1 t ) sinh ( 2 α 2 t ) − { − α 1 g ′ g + α 1 f ′ f − α 2 f g ′ − α 2 f ′ g } sin ( 2 α 1 t ) cosh ( 2 α 2 t )$
$I 5 , 2 = 1 2 [ { − ( α 1 2 − α 2 2 ) ( − g 2 + f 2 ) + 4 α 1 α 2 f g − g ′ 2 + f ′ 2 } sin ( 2 α 1 t ) sinh ( 2 α 2 t ) + { 2 α 1 α 2 ( f 2 − g 2 ) + 2 f g ( α 1 2 − α 2 2 ) − 2 f ′ g ′ } cos ( 2 α 1 t ) cosh ( 2 α 2 t ) ] − [ { α 1 f ′ g + α 1 g ′ f − α 2 f ′ f + α 2 g g ′ } sin ( 2 α 1 t ) cosh ( 2 α 2 t ) ] − [ { α 1 f ′ f − α 1 g ′ g − α 2 g ′ f − α 2 f ′ g } cos ( 2 α 1 t ) sinh ( 2 α 2 t ) ]$
The Gauss2 method is again used to integrate Equation (36) with stepsize $h = 0.01$ and n = 10,000 number of steps. The absolute error in the first integrals is calculated as before. The absolute error in integrals $I 1 , 1$, $I 1 , 2$, $I 3 , 1$ and $I 3 , 2$ is plotted in Figure 9, Figure 10, Figure 11 and Figure 12, respectively, which remains bounded for long time. Similar error behavior is obtained for $I 2 , 1$, $I 2 , 2$, $I 4 , 1$, $I 4 , 2$, $I 5 , 1$, and $I 5 , 2$. The symplectic Gauss2 method is able to preserve all first integrals obtained by performing complex symmetry analysis.

## 5. Conclusions

The first integrals of dynamical system $y ″ = − k 2 y$ were obtained via the classical Noether approach and the complex symmetry method. The later approach yields invariant energy as a particular example that is stored in both oscillators. Since these first integrals are quadratic in nature, the symplectic Runge–Kutta method, whose construction is also given in this paper, was successfully applied to the system, and numerical preservation of these first integrals was obtained. Interestingly, the numerical method presented in this paper could preserve the energy of the single oscillator as well as the energy stored in the pair of coupled oscillators that arise from the complex Noether approach. The error in the first integrals remained bounded for a long time, which would not have been possible if we have employed nonsymplectic integrators.

## Author Contributions

Conceptualization, Y.H.; formal analysis, W.I. and M.U.F.; investigation, W.I.; methodology, Y.H. and M.U.F.; project administration, Y.H.; software, Y.H. and W.I.; supervision, Y.H.; validation, M.U.F.; visualization, W.I.; writing—original draft, W.I.; writing—review and editing, Y.H.

## Funding

This research received no external funding.

## Conflicts of Interest

The authors declare that there is no conflict of interest.

## References

1. Ibragimov, N.H. Elementary Lie Group Analysis and Ordinary Differential Equations; John Wiley and Sons: Chichester, UK, 1999. [Google Scholar]
2. Lie, S. Theorie der Transformationsgruppen; Teubner: Leipzig, Germany, 1888. [Google Scholar]
3. Lie, S. Vorlesungen über Differentialgleichungen Mit Bekannten Infinitesimalen Transformationen; Teubner: Leipzig, Germany, 1891. [Google Scholar]
4. Stephani, H. Differential Equations: Their Solution Using Symmetries; Cambridge University Press: Cambridge, UK, 1989. [Google Scholar]
5. Noether, E. Invariante Variationsprobleme, Nachrichten der Akademie der Wissenschaften in Göttingen. Mathematisch-Physikalische Klasse 1918, 2, 235–257, English translation in Transport Theory and Statistical Physics 1971, 1, 186–207. [Google Scholar]
6. Ibragimov, N.H.; Kara, A.H.; Mahomed, F.M. Lie-Backlund and Noether symmetries with applications. Nonlinear Dyn. 1998, 15, 115–136. [Google Scholar] [CrossRef]
7. Leach, P.G.L. Applications of the Lie theory of extended groups in Hamiltonian mechanics: The oscillator and the Kepler problem. J. Aust. Math. Soc. 1981, 23, 173–186. [Google Scholar] [CrossRef]
8. Lutzky, M. Symmetry groups and conserved quantities for the harmonic oscillator. J. Phys. A Math. Gen. 1978, 11, 249. [Google Scholar] [CrossRef]
9. Naz, R.; Freire, I.L.; Naeem, I. Comparison of different approaches to construct first integrals for ordinary differential equations. Abstr. Appl. Anal. 2014, 2014, 1–15. [Google Scholar] [CrossRef] [PubMed]
10. Farooq, M.U.; Ali, S.; Mahomad, F.M. Two-dimensional systems that arise from the Noether classification of Lagrangians on the line. Appl. Math. Comput. 2011, 217, 6959–6973. [Google Scholar] [CrossRef] [Green Version]
11. Farooq, M.U.; Ali, S.; Qadir, A. Invariants of two-dimensional systems via complex Lagrangians with applications. Commun. Nonlinear Sci. Numer. Simul. 2011, 16, 1804–1810. [Google Scholar] [CrossRef]
12. Ali, S.; Mahomed, F.M.; Qadir, A. Complex Lie symmetries for variational problems. J. Nonlinear Math. Phys. 2008, 5, 25–35. [Google Scholar] [CrossRef]
13. Sanz-Serna, J.M.; Calvo, M.P. Symplectic numerical methods for Hamiltonian problems. J. Mod. Phys. C 1993, 4, 385–392. [Google Scholar] [CrossRef]
14. Hairer, E.; Lubich, C.; Wanner, G. Geometric Numerical Integration: Structure-Preserving Algorithms for Ordinary Differential Equations, 2nd ed.; Springer: Berlin/Heidelberg, Germany, 2005. [Google Scholar]
15. Ali, S.; Mahomed, F.M.; Qadir, A. Complex Lie symmetries for scalar second-order ordinary differential equations. Nonlinear Anal. Real World Appl. 2009, 10, 3335–3344. [Google Scholar] [CrossRef]
16. Butcher, J.C. A history of Runge-Kutta methods. Appl. Numer. Math. 1996, 20, 247–260. [Google Scholar] [CrossRef]
17. Burrage, K.; Butcher, J.C. Stability criteria for implicit Runge-Kutta methods. SIAM J. Numer. Anal. 1979, 16, 46–57. [Google Scholar] [CrossRef]
18. Cooper, G.J. Stability of Runge-Kutta methods for trajectory problems. IMA J. Numer. Anal. 1987, 7, 1–13. [Google Scholar] [CrossRef]
19. Lasagni, F.M. Canonical Runge-Kutta methods. ZAMP 1988, 39, 952–953. [Google Scholar] [CrossRef]
20. Sanz-Serna, J.M.; Calvo, M.P. Numerical Hamiltonian Problems, 1st ed.; Chapman and Hall: London, UK, 1994. [Google Scholar]
21. Sanz-Serna, J.M. Runge-Kutta schemes for Hamiltonian systems. BIT 1988, 28, 877–883. [Google Scholar] [CrossRef]
22. Sun, G. A simple way of constructing symplectic Runge-Kutta methods. J. Comput. Math. 2000, 18, 61–68. [Google Scholar]
23. Habib, Y. Long-Term Behaviour of G-Symplectic Methods. Ph.D. Thesis, The University of Auckland, Auckland, New Zealand, 2010. [Google Scholar]
Figure 1. Error in integral $I 2$ using Gauss2.
Figure 1. Error in integral $I 2$ using Gauss2.
Figure 2. Error in integral $I 2$ using symplectic Euler.
Figure 2. Error in integral $I 2$ using symplectic Euler.
Figure 3. Error in integral $I 3$ using Gauss2.
Figure 3. Error in integral $I 3$ using Gauss2.
Figure 4. Error in integral $I 3$ using symplectic Euler.
Figure 4. Error in integral $I 3$ using symplectic Euler.
Figure 5. Error in integral $I 2 , 1$.
Figure 5. Error in integral $I 2 , 1$.
Figure 6. Error in integral $I 2 , 2$
Figure 6. Error in integral $I 2 , 2$
Figure 7. Error in integral $I 4 , 1$.
Figure 7. Error in integral $I 4 , 1$.
Figure 8. Error in integral $I 4 , 2$.
Figure 8. Error in integral $I 4 , 2$.
Figure 9. Error in integral $I 1 , 1$.
Figure 9. Error in integral $I 1 , 1$.
Figure 10. Error in integral $I 1 , 2$.
Figure 10. Error in integral $I 1 , 2$.
Figure 11. Error in integral $I 3 , 1$.
Figure 11. Error in integral $I 3 , 1$.
Figure 12. Error in integral $I 3 , 2$.
Figure 12. Error in integral $I 3 , 2$.

## Share and Cite

MDPI and ACS Style

Irshad, W.; Habib, Y.; Farooq, M.U. A Complex Lie-Symmetry Approach to Calculate First Integrals and Their Numerical Preservation. Symmetry 2019, 11, 11. https://doi.org/10.3390/sym11010011

AMA Style

Irshad W, Habib Y, Farooq MU. A Complex Lie-Symmetry Approach to Calculate First Integrals and Their Numerical Preservation. Symmetry. 2019; 11(1):11. https://doi.org/10.3390/sym11010011

Chicago/Turabian Style

Irshad, Wajeeha, Yousaf Habib, and Muhammad Umar Farooq. 2019. "A Complex Lie-Symmetry Approach to Calculate First Integrals and Their Numerical Preservation" Symmetry 11, no. 1: 11. https://doi.org/10.3390/sym11010011

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.