MPE Mathematical Problems in Engineering 1563-5147 1024-123X Hindawi Publishing Corporation 10.1155/2015/438694 438694 Research Article On Some Properties and Symmetries of the 5-Dimensional Lorenz System Lăzureanu Cristian Bînzar Tudor Wang Yan-Wu Department of Mathematics Politehnica University of Timişoara Piaţa Victoriei No. 2 300006 Timişoara Romania upt.ro 2015 29102015 2015 11 08 2015 03 10 2015 05 10 2015 29102015 2015 Copyright © 2015 Cristian Lăzureanu and Tudor Bînzar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The 5-dimensional Lorenz system for the gravity-wave activity is considered. Some stability problems and the existence of periodic orbits are studied. Also, a symplectic realization and some symmetries are given.

1. Introduction

The importance of the 5-dimensional Lorenz system  in the study of geophysical fluid dynamics is well known. This system describes coupled Rossby waves and gravity waves. It was mainly investigated from the existence of a slow manifold point of view . Among other studies regarding 5-dimensional Lorenz system we mention Hamiltonian structure , chaotic behaviour , and analytic integrability .

According to , the 5-dimensional Lorenz system has at most three functionally independent global analytic first integrals. We mention that two first integrals are known . It raises the following question: how can the third first integral be determined, provided that it exists? A possible answer is given by the connection between symmetries and the existence of conservative laws . Our main purpose is to try to determine the third first integral using symmetries. This attempt was successful in the case of 5-dimensional Maxwell-Bloch equations with the rotating wave approximation . “Intuitively speaking, a symmetry is a transformation of an object leaving this object invariant” . In our case, a transformation means a vector field and an object means a differential equation. Recently, this field is widely investigated. We refer to some new progress .

In our paper, the constants of motion of the 5-dimensional Lorenz system are used to study some stability problems and the existence of periodic orbits. “The stability of an orbit of a dynamical system characterizes whether nearby (i.e., perturbed) orbits will remain in a neighborhood of that orbit or be repelled away from it” . Also, with the aid of these constants of motion, a symplectic realization and a Lagrangian formulation are given. In the last part of our work some symmetries are pointed out.

2. Stability and Periodic Orbits

We consider the 5-dimensional Lorenz system : (1) x ˙ 1 = - x 2 x 3 + b x 2 x 5 , x ˙ 2 = x 1 x 3 - b x 1 x 5 , x ˙ 3 = - x 1 x 2 , x ˙ 4 = - x 5 , x ˙ 5 = x 4 + b x 1 x 2 , where b R .

Recall that, for system (1), the functions H , C C ( R 5 , R ) , (2) H x 1 , x 2 , x 3 , x 4 , x 5 = 1 2 x 1 2 + 2 x 2 2 + x 3 2 + x 4 2 + x 5 2 , (3) C x 1 , x 2 , x 3 , x 4 , x 5 = 1 2 x 1 2 + x 2 2 are constants of motion. The functions H and C are linearly related to analogs of the energy and, respectively, enstrophy of the nine-component “primitive equations” model introduced by Lorenz [1, 8].

Considering the matrix formulation of the Poisson bracket { · , · } , given in coordinates by (4) π = 0 0 - x 2 0 b x 2 0 0 x 1 0 - b x 1 x 2 - x 1 0 0 0 0 0 0 0 - 1 - b x 2 b x 1 0 1 0 , system (1) has the Hamiltonian form : (5) x ˙ = x , H , where the Hamiltonian H is given by (2). Hence ( R 5 , π , X H ) is a Hamilton-Poisson realization of dynamics (1), where (6) X H = - x 2 x 3 + b x 2 x 5 , x 1 x 3 - b x 1 x 5 , - x 1 x 2 , - x 5 , x 4 + b x 1 x 2 . It is easy to see that the function C is a Casimir for the above Poisson bracket.

In the following we study the stability of system (1).

The equilibrium states of system (1) are given as the union of the following families: (7) E 1 = M , 0,0 , 0,0 M R , E 2 = 0 , M , 0,0 , 0 M R , E 3 = 0,0 , M , 0,0 M R . Let e M 1 = ( M , 0,0 , 0,0 ) E 1 . Considering the function L C ( R 5 , R ) , (8) L x 1 , x 2 , x 3 , x 4 , x 5 = x 2 2 + x 3 2 + x 4 2 + x 5 2 + x 1 2 + x 2 2 - M 2 2 , we have (9) L e M 1 = 0 , L x 1 , x 2 , x 3 , x 4 , x 5 > 0 , x 1 , x 2 , x 3 , x 4 , x 5 V e M 1 , L ˙ x 1 , x 2 , x 3 , x 4 , x 5 = 2 x 2 x ˙ 2 + 2 x 3 x ˙ 3 + 2 x 4 x ˙ 4 + 2 x 5 x ˙ 5 + 4 x 1 2 + x 2 2 - M 2 x 1 x ˙ 1 + x 2 x ˙ 2 = 0 , x 1 , x 2 , x 3 , x 4 , x 5 V , for  some  neighbourhood   V   of   e M 1 . By [19, 20], we deduce that all the equilibrium states from the family E 1 are nonlinearly stable.

The characteristic polynomial associated with the linear part of system (1) at the equilibrium e M 2 = ( 0 , M , 0,0 , 0 ) , M 0 , is given by (10) p e M 2 λ = - λ λ 4 - M 2 + b 2 M 2 - 1 λ 2 - M 2 . We notice that a root of p E M 2 is strictly positive, whence e M 2 is an unstable equilibrium state. Therefore, all the equilibrium states from the family E 2 are unstable.

Let e M 3 E 3 . The roots of the characteristic polynomial associated with the linear part of system (1) at e M 3 are (11) λ 1 = 0 , λ 2,3 = ± i , λ 4,5 = ± i M . Hence all the equilibrium states from the family E 3 are spectrally stable.

Now, we study the existence of periodic orbits of system (1) around the equilibrium states from the family E 1 { ( 0,0 , 0,0 , 0 ) } .

Since the eigenvalues of the linear part of system (1) at the equilibrium e M 1 = ( M , 0,0 , 0,0 ) , M 0 , are (12) λ 1 = 0 , λ 2,3 = ± i - y 1 , λ 4,5 = ± i - y 2 , where y 1 and y 2 are the roots of the equation (13) y 2 + M 2 + b 2 M 2 + 1 y + M 2 = 0 , we apply Theorem 2.1 from . The eigenspace corresponding to the eigenvalue λ 1 = 0 has one dimension. Taking the constant of motion I : R 5 R , (14) I x 1 , x 2 , x 3 , x 4 , x 5 = x 2 2 + x 3 2 + x 4 2 + x 5 2 , it follows that (15) d I e M 1 = 0 , d 2 I e M 1 W × W > 0 , where (16) W = ker d C e M 1 = Span R 0,1 , 0,0 , 0 , 0,0 , 1,0 , 0 , 0,0 , 0,1 , 0 , 0,0 , 0,0 , 1 . Therefore, for each sufficiently small ɛ R + , any integral hypersurface (17) Σ ɛ e M 1 : x 2 2 + x 3 2 + x 4 2 + x 5 2 = ɛ 2 contains at least one periodic orbit of system (1) whose period is close to 2 π / - y 1 and at least one periodic orbit of system (1) whose period is close to 2 π / - y 2 .

In the case of the equilibrium states from E 3 , we cannot apply the above method. On the other hand the dynamics of system (1) are carried out at the intersection of the hypersurfaces (18) H x 1 , x 2 , x 3 , x 4 , x 5 = H 0,0 , M , 0,0 , C x 1 , x 2 , x 3 , x 4 , x 5 = C 0,0 , M , 0,0 ; that is, (19) x 3 2 + x 4 2 + x 5 2 = M 2 , x 1 2 + x 2 2 = 0 . Then the solution of system (1) is (20) x 1 = 0 , x 2 = 0 , x 3 = ± M 2 - k 1 2 - k 2 2 , x 4 = k 1 cos t + k 2 sin t , x 5 = k 1 sin t - k 2 cos t , where k 1 , k 2 R , k 1 2 + k 2 2 M 2 . We remark that (20) represents periodic orbits around equilibrium state ( 0,0 , M , 0,0 ) , M R (see Figure 1).

Periodic orbits around equilibrium state ( 0,0 , M , 0,0 ) (3D view in the x 3 - x 4 - x 5 space).

3. Symplectic Realization and Symmetries

First result shows that system (1) can be regarded as a Hamiltonian mechanical system.

Theorem 1.

The Hamilton-Poisson mechanical system ( R 5 , π , X H ) has a full symplectic realization ( T R 3 R 6 , ω , X H ~ ) , where (21) ω = d p 1 d q 1 + d p 2 d q 2 + d p 3 d q 3 , H ~ = 1 2 p 1 2 + p 2 2 + p 3 2 + 1 2 q 2 - b p 1 2 + 1 2 p 3 2 sin 2 q 1 , and the corresponding Hamiltonian vector field is as follows: (22) X H ~ = 1 + b 2 p 1 - b q 2 q 1 + p 2 q 2 + p 3 + p 3 sin 2 q 1 q 3 - p 3 2 sin q 1 cos q 1 p 1 + b p 1 - q 2 p 2 .

Proof.

Using the Hamiltonian H ~ one obtains the corresponding Hamilton’s equations: (23) q ˙ 1 = 1 + b 2 p 1 - b q 2 , q ˙ 2 = p 2 , q ˙ 3 = p 3 + p 3 sin 2 q 1 , p ˙ 1 = - p 3 2 sin q 1 cos q 1 , p ˙ 2 = b p 1 - q 2 , p ˙ 3 = 0 . We consider the mapping φ : R 6 R 5 , (24) φ q 1 , q 2 , q 3 , p 1 , p 2 , p 3 = p 3 cos q 1 , p 3 sin q 1 , p 1 , p 2 , q 2 - b p 1 = x 1 , x 2 , x 3 , x 4 , x 5 . Using the standard symplectic bracket (25) f , g ω = i = 1 3 f q i · g p i - f p i · g q i , one gets (26) x 1 , x 2 ω = 0 , x 1 , x 3 ω = - x 2 , x 1 , x 4 ω = 0 , x 1 , x 5 ω = b x 2 , x 2 , x 3 ω = x 1 , x 2 , x 4 ω = 0 , x 2 , x 5 ω = - b x 1 , x 3 , x 4 ω = 0 , x 3 , x 5 ω = 0 , x 4 , x 5 ω = - 1 . Hence the canonical structure { · , · } ω is mapped onto the Poisson structure π .

Taking into account relations (23), we have (27) x ˙ 1 = - p 3 sin q 1 · q ˙ 1 = - x 2 x 3 + b x 2 x 5 , x ˙ 2 = p 3 cos q 1 · q ˙ 1 = x 1 x 3 - b x 1 x 5 , x ˙ 3 = p ˙ 1 = - x 1 x 2 , x ˙ 4 = p ˙ 2 = - x 5 , x ˙ 5 = q ˙ 2 - b p ˙ 1 = x 4 + b x 1 x 2 . Therefore the Hamiltonian vector field X H ~ is mapped onto the Hamiltonian vector field X H . Moreover φ is a surjective submersion and H φ = H ~ , which finishes the proof.

Denoting C ~ C φ , it follows that C ~ = p 3 .

The next result states that system (23) can be written in Lagrangian formalism.

Theorem 2.

System (23) has the form (28) q ¨ 1 + 1 + b 2 sin q 1 cos q 1 1 + sin 2 q 1 2 q ˙ 3 2 + b q ˙ 2 = 0 , q ¨ 2 - b 1 + b 2 q ˙ 1 + 1 1 + b 2 q 2 = 0 , q ¨ 3 - 2 sin q 1 cos q 1 1 + sin 2 q 1 q ˙ 1 q ˙ 3 = 0 , on the tangent bundle T R 3 . Also, system (28) represents the Euler-Lagrange equations generated by the Lagrangian (29) L = 1 2 1 + b 2 q ˙ 1 2 + 1 2 q ˙ 2 2 + 1 2 1 + sin 2 q 1 q ˙ 3 2 + b 1 + b 2 q ˙ 1 q 2 - 1 2 1 + b 2 q 2 2 .

Proof.

By Hamilton’s equations (23) we obtain (30) p 1 = 1 1 + b 2 q ˙ 1 + b 1 + b 2 q 2 p 2 = q ˙ 2 p 3 = 1 1 + sin 2 q 1 q ˙ 3 , whence (31) p ˙ 1 = 1 1 + b 2 q ¨ 1 + b 1 + b 2 q ˙ 2 , p ˙ 2 = q ¨ 2 , p ˙ 3 = - 2 sin q 1 cos q 1 1 + sin 2 q 1 2 q ˙ 1 q ˙ 3 + 1 1 + sin 2 q 1 q ¨ 3 . Substituting p 1 , p 2 , p 3 , p ˙ 1 , p ˙ 2 , p ˙ 3 into (23), one gets (28). For the Lagrangian L given by (29), the Euler-Lagrange equations, (32) d d t L q ˙ i - L q i = 0 , i 1,2 , 3 , have the form (28). The relation between the Hamiltonian H ~ and the Lagrangian L , (33) H ~ = i = 1 3 p i q ˙ i - L , where (34) p i = L q ˙ i , i 1,2 , 3 , follows by relations (23).

For details about Lagrangian and Hamiltonian formalism see, for example, [22, 23].

In the sequel we study the Lie-point symmetries for Euler-Lagrange equations (28).

We recall that a vector field (35) v = ξ q 1 , q 2 , q 3 , t t + η 1 q 1 , q 2 , q 3 , t q 1 + η 2 q 1 , q 2 , q 3 , t q 2 + η 3 q 1 , q 2 , q 3 , t q 3 is a Lie-point symmetry for Euler-Lagrange equations if the action of its second prolongation on these equations vanishes. For more details about symmetries see, for example, .

Applying the second prolongation of v , (36) pr 2 v = v + i = 1 3 η ˙ i - ξ ˙ q ˙ i q ˙ i + i = 1 3 η ¨ i - ξ ¨ q ˙ i - 2 ξ ˙ q ¨ i q ¨ i , on (28) one obtains (37) η ¨ 1 - ξ ¨ q ˙ 1 - 2 q ¨ 1 ξ ˙ + 1 + b 2 1 - 5 sin 2 q 1 + 2 sin 4 q 1 1 + sin 2 q 1 3 q ˙ 3 2 η 1 + b η ˙ 2 - ξ ˙ q ˙ 2 + 2 1 + b 2 sin q 1 cos q 1 1 + sin 2 q 1 2 η ˙ 3 - ξ ˙ q ˙ 3 q ˙ 3 = 0 , η ¨ 2 - ξ ¨ q ˙ 2 - 2 q ¨ 2 ξ ˙ + 1 1 + b 2 η 2 - b 1 + b 2 η ˙ 1 - ξ ˙ q ˙ 1 = 0 , η ¨ 3 - ξ ¨ q ˙ 3 - 2 q ¨ 3 ξ ˙ + 2 3 sin 2 q 1 - 1 1 + sin 2 q 1 2 η 1 q ˙ 1 q ˙ 3 - 2 sin q 1 cos q 1 1 + sin 2 q 1 η ˙ 1 - ξ ˙ q ˙ 1 q ˙ 3 - 2 sin q 1 cos q 1 1 + sin 2 q 1 η ˙ 3 - ξ ˙ q ˙ 3 q ˙ 1 = 0 . The resulting equations obtained by expanding ξ ˙ , ξ ¨ , η ˙ 1 , η ¨ 1 , η ˙ 2 , η ¨ 2 , η ˙ 3 , η ¨ 3 and replacing q ¨ 1 , q ¨ 2 , and q ¨ 3 must be satisfied identically in t , q 1 , q 2 , q 3 , q ˙ 1 , q ˙ 2 , q ˙ 3 , which are all independent variables.

In the case b 0 , it follows that (38) ξ q 1 = ξ q 2 = ξ q 3 = 0 , η 1 , t = η 1 , q 2 = η 1 , q 3 = 0 , η 1 , q 1 q 1 = 0 , η 2 , q 1 = η 2 , q 3 = 0 , η 2 , q 2 q 2 = 0 , η 3 , t = η 3 , q 1 = η 3 , q 2 = 0 , η 3 , q 3 q 3 = 0 , η 2 + 1 + b 2 η 2 , t t - q 2 η 2 , q 2 + 2 q 2 ξ t = 0 , η 1 , q 1 + ξ t - η 2 , q 2 = 0 , 2 η 2 , t q 2 - ξ t t = 0 , b η 2 , t = 0 , b η 2 , q 2 + ξ t - η 1 , q 1 = 0 , 1 - 5 sin 2 q 1 + 2 sin 4 q 1 1 + sin 2 q 1 η 1 + sin q 1 cos q 1 2 η 3 , q 3 - η 1 , q 1 = 0 , 6 sin 2 q 1 - 2 1 + sin 2 q 1 η 1 - 2 sin q 1 cos q 1 η 1 , q 1 = 0 . The last relation implies η 1 = 0 . It results in (39) ξ = α , η 1 = 0 , η 2 = 0 , η 3 = β , α , β R .

In the case b = 0 , it follows that (40) ξ q 1 = ξ q 2 = ξ q 3 = 0 , η 1 = 0 , η 2 , q 1 = η 2 , q 3 = 0 , η 2 , q 2 q 2 = 0 , η 2 , t q 2 = 0 , η 3 , t = η 3 , q 1 = η 3 , q 2 = η 3 , q 3 = 0 , η 2 + η 2 , t t - q 2 η 2 , q 2 + 2 q 2 ξ t = 0 . It results in (41) ξ = a t + b , η 2 = c q 2 + f t , a , b , c R . Then a = 0 and f ( t ) + f ( t ) = 0 . Therefore (42) ξ = α , η 1 = 0 , η 2 = γ q 2 + δ cos t + θ sin t , η 3 = β , α , β , γ , δ , θ R .

We can conclude the following result.

Theorem 3.

The symmetries of (28) are given by (43) v = α t + β q 3 , where α , β R , in the case b R , respectively, and (44) u = α t + γ q 2 + δ cos t + θ sin t q 2 + β q 3 , where α , β , γ , δ , θ R , in the case b = 0 .

Remark 4.

Let γ = δ = θ = 0 . Denoting v 1 = / t and v 2 = / q 3 , it follows that v 1 , v 2 are variational symmetries. Moreover,

(i) for β = 0 and α 0 , we have v = α v 1 that represents the time translation symmetry which generates the conservation of energy H ~ ;

(ii) for α = 0 and β 0 , we have v = β v 2 that represents a translation in the cyclic q 3 direction which is related to the conservation of p 3 .

We notice that the vector field u leads to the vector field (45) X = α t + γ x 4 - δ sin t + θ cos t x 4 + γ x 5 + δ cos t + θ sin t x 5 . Also, we can consider the vector field (46) Y = α t + γ x 5 - δ sin t + θ cos t x 4 + - γ x 4 + δ cos t + θ sin t x 5 .

The last result furnishes some symmetries of system (1) in the case b = 0 .

Proposition 5.

The vector field X given by (45) is a Lie-point symmetry of system (1) in the case b = 0 . Also, if δ = θ = 0 , then X is a symmetry of system (1) in the case b = 0 . Moreover, the vector field Y given by (46) has the same properties.

Proof.

It is easy to see that the action of the first prolongation of X on (1) in the case b = 0 vanishes. Therefore X is a Lie-point symmetry.

Considering δ = θ = 0 , it immediately follows that (47) X t + X , X f = 0 , where (48) X f = - x 2 x 3 , x 1 x 3 , - x 1 x 3 , - x 5 , x 4 , whence X is a symmetry of system (1) in the case b = 0 .

4. Conclusions

In this paper the 5-dimensional Lorenz system is considered. This is a system of five differential equations which couples the Rossby waves and gravity waves. In Section 2 some stability problems and the existence of periodic orbits are studied. The equilibrium states of considered system are given as the union of three families of points. For one of these families, all the equilibria are spectrally stable, but it remains an open problem to establish if these equilibria are nonlinearly stable. In the third part of the paper a symplectic realization and the corresponding Lagrangian formulation are given. In the last part of our work, some symmetries of the 5-dimensional Lorenz system are studied. Knowing the connection between symmetries and conservative laws, we tried to determine a third first integral of the considered system, provided that it exists.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Lorenz E. N. On the existence of a slow manifold Journal of the Atmospheric Sciences 1986 43 15 1547 1558 10.1175/1520-0469(1986)04360;1547:oteoas62;2.0.co;2 Lorenz E. N. Krishnamurthy V. On the nonexistence of a slow manifold Journal of the Atmospheric Sciences 1987 44 20 2940 2950 10.1175/1520-0469(1987)04460;2940:otnoas62;2.0.co;2 Boyd J. P. The slow manifold of a five-mode model Journal of the Atmospheric Sciences 1994 51 8 1057 1064 10.1175/1520-0469(1994)051<1057:TSMOAF>2.0.CO;2 2-s2.0-0028607751 Camassa R. Tin S.-K. The global geometry of the slow manifold in the Lorenz-Krishnamurthy model Journal of the Atmospheric Sciences 1996 53 22 3251 3264 10.1175/1520-0469(1996)053<3251:TGGOTS>2.0.CO;2 2-s2.0-0030357843 Ginoux J.-M. The slow invariant manifold of the Lorenz-Krishnamurty model Qualitative Theory of Dynamical Systems 2014 13 1 19 37 10.1007/s12346-013-0104-6 2-s2.0-84898412295 Bokhove O. On Hamiltonian balanced models Proceedings of the 9th Conference on Atmospheric and Oceanic Waves and Stability 1993 San Antonio, Tex, USA American Meteorological Society 367 368 Camassa R. On the geometry of an atmospheric slow manifold Physica D: Nonlinear Phenomena 1995 84 3-4 357 397 10.1016/0167-2789(94)00239-m 2-s2.0-0000025724 Bokhove O. Shepherd T. G. On Hamiltonian balanced dynamics and the slowest invariant manifold Journal of the Atmospheric Sciences 1996 53 2 276 297 10.1175/1520-0469(1996)053<0276:OHBDAT>2.0.CO;2 2-s2.0-0029656559 Birtea P. Puta M. Ratiu T. S. Tudoran R. A short proof of chaos in an atmospheric system Physics Letters A 2002 300 2-3 189 191 10.1016/s0375-9601(02)00828-9 2-s2.0-0037194338 Llibre J. Saghin R. Zhang X. On the analytic integrability of the 5-dimensional lorenz system for the gravity-wave activity Proceedings of the American Mathematical Society 2014 142 2 531 537 2-s2.0-84889014989 Noether E. Invariante variations probleme Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen, Mathematisch-Physikalische Klasse 1918 1 235 257 English translation: Transport Theory and Statistical Physics, vol. 1, pp. 186–207, 1971 Caşu I. Symmetries of the Maxwell-Bloch equations with the rotating wave approximation Regular and Chaotic Dynamics 2014 19 5 548 555 10.1134/S1560354714050037 ZBL1308.70018 2-s2.0-84910131758 Bihlo A. Symmetry methods in the atmospheric sciences [Ph.D. thesis] 2010 Universitat Wien Gandarias M. L. Bruzón M. S. Rosa M. Symmetries and conservation laws for some compacton equation Mathematical Problems in Engineering 2015 2015 6 430823 10.1155/2015/430823 Gray R. J. The Lie point symmetry generators admitted by systems of linear differential equations Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences 2014 470 2166 12 10.1098/rspa.2013.0779 2-s2.0-84899829080 Naz R. Naeem I. Mahomed F. M. A partial lagrangian approach to mathematical models of epidemiology Mathematical Problems in Engineering 2015 2015 11 602915 10.1155/2015/602915 Cao Z. Lin Y. Lie point symmetries, conservation laws, and solutions of a space dependent reaction-diffusion equation Applied Mathematics and Computation 2014 248 1 386 398 10.1016/j.amc.2014.09.093 2-s2.0-84908377715 Holmes P. Shea-Brown E. T. Stability Scholarpedia 2006 1 10, article 1838 10.4249/scholarpedia.1838 Lyapunov A. M. Problème général de la stabilité du mouvement Annales de la Faculté des Sciences de Toulouse 1907 9 203 474 Kharkov 1892 Lyapunov A. M. Problème Général de la Stabilité du Mouvement 1949 17 Princeton, NJ, USA Princeton University Press Birtea P. Puta M. Tudoran R. M. Periodic orbits in the case of a zero eigenvalue Comptes Rendus Mathematique 2007 344 12 779 784 10.1016/j.crma.2007.05.003 ZBL1131.34034 2-s2.0-34447322019 Libermann P. Marle C.-M. Symplectic Geometry and Analytical Mechanics 1987 Dordrecht, The Netherlands D. Reidel Publishing Marsden J. E. Raţiu T. S. Introduction to Mechanics and Symmetry 1999 17 2nd Berlin, Germany Springer Texts in Applied Mathematics 10.1007/978-0-387-21792-5 MR1723696 Bluman G. W. Kumei S. Symmetries and Differential Equations 1989 81 New York, NY, USA Springer Applied Mathematical Sciences 10.1007/978-1-4757-4307-4 MR1006433 Olver P. J. Applications of Lie Groups to Differential Equations 1986 New York, NY, USA Springer Lăzureanu C. Bînzar T. Symmetries of some classes of dynamical systems Journal of Nonlinear Mathematical Physics 2015 22 2 265 274 10.1080/14029251.2015.1033237