<?xml version="1.0" encoding="utf-8"?>
<?xml-stylesheet type="text/css" href="math.css"?>
<!DOCTYPE article [<!ATTLIST article xmlns CDATA "http://xml-maiden.com"><!ATTLIST a xmlns CDATA "http://www.w3.org/1999/xhtml">]>
<article>
<title>Non-Noether symmetries in singular dynamical systems</title>
<author>George Chavchanidze</author>
<affiliation>Department of Theoretical Physics,
A. Razmadze Institute of Mathematics,
1 Aleksidze Street, Tbilisi 0193, Georgia</affiliation>
<abstract>In the present paper geometric aspects of relationship
between non-Noether symmetries and conservation laws in Hamiltonian
systems is discussed. Case of irregular/constrained dynamical systems
on presymplectic and Poisson manifolds is considered.</abstract>
<keywords>Non-Noether symmetry; Conservation laws; Constrained dynamics;</keywords>
<msc> 70H33, 70H06, 53Z05</msc>
<reference>Georgian Math. J. 10 (2003) 057-061</reference>

<header xml:id="intro">Introduction</header>
<paragraph>
Noether's theorem associates conservation laws with particular continuous symmetries of
the Lagrangian. According to the Hojman's theorem <a href="#r1">[1]</a>-<a href="#r3">[3]</a> 
there exists the definite correspondence between
non-Noether symmetries and conserved quantities. In 1998 M. Lutzky showed that several integrals of
motion might correspond to a single one-parameter group of non-Noether transformations
<a href="#r4">[4]</a>. In the present paper, the extension of Hojman-Lutzky theorem to singular dynamical systems is considered.
</paragraph>
<paragraph>
First of all let us recall some basic knowledge of description of the regular dynamical systems
(see, e. g. <a href="#r5">[5]</a>).
In this case time evolution is governed by Hamilton's equation
<formula>
i<sub>X<sub>h</sub></sub>ω + dh = 0,
</formula>
where <math>ω</math> is the closed
(<math>dω = 0</math>) and non-degenerate
(<math>i<sub>X</sub>ω = 0  ⇒ X = 0</math>) 2-form,
<math>h</math> is the Hamiltonian and
<math>i<sub>X</sub>ω</math> denotes contraction of
<math>X</math> with <math>ω</math>.
Since <math>ω</math> is non-degenerate, this gives rise to an isomorphism between the vector
fields and 1-forms given by <math>i<sub>X</sub>ω + α= 0</math>.
The vector field is said to be Hamiltonian if it corresponds to exact form
<formula>
i<sub>X<sub>f</sub></sub>ω + df = 0.
</formula>
The Poisson bracket is defined as follows:
<formula>
{f , g} = X<sub>f</sub> g = − X<sub>g</sub> f = i<sub>X<sub>f</sub></sub>
i<sub>X<sub>g</sub></sub>ω.
</formula>
By introducing a bivector field <math>W</math> satisfying
<formula>
i<sub>X</sub>i<sub>Y</sub>ω = i<sub>W</sub> i<sub>X</sub>ω ∧ i<sub>Y</sub>ω,
</formula>
Poisson bracket can be rewritten as
<formula>
{f , g} = i<sub>W</sub> df ∧ dg.
</formula>
It's easy to show that
<formula xml:id="e6">
i<sub>X</sub>i<sub>Y</sub>L<sub>Z</sub>ω =
i<sub>[Z,W]</sub> i<sub>X</sub>ω ∧ i<sub>Y</sub>ω,
</formula>
where the bracket <math>[ · , · ]</math> is actually a supercommutator,
for an arbitrary bivector field
<math> W = <under><ope>∑</ope><sat>s</sat></under> V<sup>s</sup> ∧ U<sup>s</sup> </math> we have
<formula>
[X,W] = <under><ope>∑</ope><sat>s</sat></under>[X,V<sup>s</sup>] ∧ U<sup>s</sup>
+ <under><ope>∑</ope><sat>s</sat></under>V<sup>s</sup> ∧ [X,U<sup>s</sup>]
</formula>
Equation <a href="#e6">(6)</a> is based on the following useful property of the Lie derivative
<formula>
L<sub>X</sub>i<sub>W</sub>ω = i<sub>[X,W]</sub>ω +
i<sub>W</sub>L<sub>X</sub>ω.
</formula>
Indeed, for an arbitrary bivector field
<math>W = <under><ope>∑</ope><sat>s</sat></under> V<sup>s</sup> ∧ U<sup>s</sup> </math> we have
<formula>
L<sub>X</sub>i<sub>W</sub>ω = L<sub>X</sub><under><ope>∑</ope><sat>s</sat></under>i<sub>V<sup>s</sup> ∧ U<sup>s</sup></sub>ω =
L<sub>X</sub><under><ope>∑</ope><sat>s</sat></under> i<sub>U<sup>s</sup></sub>i<sub>V<sup>s</sup></sub>ω<line />
= <under><ope>∑</ope><sat>s</sat></under> i<sub>[X,U<sup>s</sup>]</sub>i<sub>V<sup>s</sup></sub>ω +
<under><ope>∑</ope><sat>s</sat></under> i<sub>U<sup>s</sup></sub>i<sub>[X,V<sup>s</sup>]</sub>ω +
<under><ope>∑</ope><sat>s</sat></under>i<sub>U<sup>s</sup></sub>i<sub>V<sup>s</sup></sub>L<sub>X</sub>ω =
i<sub>[X,W]</sub>ω + i<sub>W</sub>L<sub>X</sub>ω
</formula>
where <math>L<sub>Z</sub></math> denotes the Lie derivative along the vector field <math>Z</math>.
According to Liouville's theorem Hamiltonian vector field
preserves <math>ω</math>
<formula>
L<sub>X<sub>f</sub></sub>ω = 0;
</formula>
therefore it commutes with <math>W</math>:
<formula>
[X<sub>f</sub> ,W] = 0.
</formula>
In the local coordinates <math> z<sub>s</sub> </math> where
<math>ω = <under><ope>∑</ope><sat>rs</sat></under>ω<sup>rs</sup>dz<sub>r</sub> ∧ z<sub>s</sub></math> bivector field
<math>W</math> has the following form
<math>W = <under><ope>∑</ope><sat>rs</sat></under>W<sup>rs</sup><fraction><num>∂</num><den>∂z<sub>r</sub></den></fraction> ∧ <fraction><num>∂</num><den>∂z<sub>s</sub></den></fraction></math> where
<math>W<sup>rs</sup></math> is matrix inverted to <math>ω<sup>rs</sup></math>.
</paragraph>

<header xml:id="regular">Case of regular Lagrangian systems</header>

<paragraph>
We can say that a group of transformations
<math>g(z) = e<sup>zL<sub>E</sub></sup></math> generated by the vector
field <math>E</math> maps the space of solutions of equation onto itself if
<formula>
i<sub>X<sub>h</sub></sub>g<sub>*</sub>(ω) + g<sub>*</sub>(dh) = 0
</formula>
For <math>X<sub>h</sub></math> satisfying
<formula>
i<sub>X<sub>h</sub></sub>ω + dh = 0
</formula>
Hamilton's equation.
It's easy to show that the vector field <math>E</math> should satisfy
<math>[E , X<sub>h</sub>] = 0</math>
Indeed,
<formula>
i<sub>X<sub>h</sub></sub>L<sub>E</sub>ω + dL<sub>E</sub>h =
L<sub>E</sub>(i<sub>X<sub>h</sub></sub>ω + dh) = 0
</formula>
since <math>[E,X<sub>h</sub>] = 0</math>. 
When <math>E</math> is not Hamiltonian,
the group of transformations <math>g(z) = e<sup>zL<sub>E</sub></sup></math> is non-Noether
symmetry (in a sense that it maps solutions onto solutions but does not preserve action).
</paragraph>
<theorem>
 (Lutzky, 1998) If the vector field <math>E</math> generates non-Noether symmetry, 
then the following functions are constant along solutions:
<formula xml:id="e15">
I<sup>(k)</sup> = i<sub>W<sup>k</sup></sub> ω<sub>E</sub><sup>k</sup>        k = 1...n,
</formula>
where <math>W<sup>k</sup></math> and <math>ω<sub>E</sub><sup>k</sup></math> are outer
powers of <math>W</math> and <math>L<sub>E</sub>ω</math>.
</theorem>
<proof> 
We have to prove that <math>I<sup>(k)</sup></math> is constant along
the flow generated by the Hamiltonian. In other words, we should find that
<math>L<sub>X<sub>h</sub></sub>I<sup>(k)</sup> = 0</math> is
fulfilled. Let us consider
<math>L<sub>X<sub>h</sub></sub>I<sup>(1)</sup></math>
<formula>
L<sub>X<sub>h</sub></sub>I<sup>(1)</sup>
= L<sub>X<sub>h</sub></sub>(i<sub>W</sub>ω<sub>E</sub>) =
i<sub>[X<sub>h</sub> , W]</sub>ω<sub>E</sub>
+ i<sub>W</sub>L<sub>X<sub>h</sub></sub>ω<sub>E</sub>,
</formula>
where according to Liouville's theorem both terms
<math>[X<sub>h</sub> , W] = 0</math> and
<formula>
i<sub>W</sub>L<sub>X<sub>h</sub></sub>L<sub>E</sub>ω =
i<sub>W</sub>L<sub>E</sub>L<sub>X<sub>h</sub></sub>ω =
0</formula>
 since <math>[E , X<sub>h</sub>] = 0</math> and 
 <math>L<sub>X<sub>h</sub></sub>ω = 0</math> vanish.
In the same manner one can verify that
<math>L<sub>X<sub>h</sub></sub>I<sup>(k)</sup> = 0</math>
</proof>
<remark> 
Theorem is valid for a larger class of generators <math>E</math> .
Namely, if <math>[E , X<sub>h</sub>] = X<sub>f</sub></math> where <math>X<sub>f</sub></math> is
an arbitrary Hamiltonian vector field, then <math>I<sup>(k)</sup></math> is still conserved. Such a
symmetries map the solutions of the equation
<math>i<sub>X<sub>h</sub></sub>ω + dh = 0</math>
on solutions of
<formula>
i<sub>X<sub>h</sub></sub>g<sub>*</sub>(ω) +
d(g<sub>*</sub>h + f) = 0</formula>
</remark>
<remark> 
 Discrete non-Noether symmetries give rise to the conservation of
<math>I<sup>(k)</sup> = i<sub>W<sup>k</sup></sub>g<sub>*</sub>(ω)<sup>k</sup></math>
where <math>g<sub>*</sub>(ω)</math> is transformed <math>ω</math>.
</remark>
<remark> 
 If <math>I<sup>(k)</sup></math> is a set of conserved quantities
associated with <math>E</math> and <math>f</math> is any conserved quantity, then the set of functions
<math>{I<sup>(k)</sup> , f} </math>
(which due to the Poisson theorem are integrals of motion) is associated with
<math>[X<sub>h</sub> , E]</math>. Namely it is easy to show by taking the Lie
derivative of <a href="#e15">(15)</a> along vector field <math> E</math> that
<formula>
{I<sup>(k)</sup> , f} = i<sub>W<sup>k</sup></sub>ω<sup>k</sup><sub>[X<sub>f</sub> , E]</sub></formula>
 is fulfilled.
As a result conserved quantities associated with Non-Noether symmetries form Lie algebra under
the Poisson bracket.
</remark>
<remark> 
If generator of symmetry satisfies Yang-Baxter equation
<math>[[E[E , W]]W] = 0</math> Lutzky's conservation laws are in involution [7]
<math>{Y<sup>(l)</sup> , Y<sup>(k)</sup>} = 0</math>
</remark>

<header xml:id="irregular">Case of irregular Lagrangian systems</header>

<paragraph>
The singular Lagrangian (Lagrangian with vanishing Hessian) leads to degenerate 2-form
<math>ω</math> and we no longer have isomorphism between vector fields and 1-forms.
Since there exists a set of "null vectors" <math>u<sub>s</sub></math> such that
<math>i<sub>u<sub>s</sub></sub>ω = 0       s = 1,2 ... n − rank(ω),</math>
every Hamiltonian vector field is
defined up to linear combination of vectors <math>u<sub>s</sub></math>. By identifying <math>X<sub>f</sub></math>
with <math>X<sub>f</sub> + <under><ope>∑</ope><sat>s</sat></under>C<sub>s</sub>u<sub>s</sub>,</math> we can introduce equivalence class
<math> X<sub>f</sub><sup>∗</sup></math> (then all <math>u<sub>s</sub></math> belong to
<math>0<sup>∗</sup></math> ).
The bivector field <math>W</math> is also far from being unique, but if
<math>W<sub>1</sub></math> and <math>W<sub>2</sub></math> both satisfy
<formula>
i<sub>X</sub>i<sub>Y</sub> ω =
i<sub>W<sub>1,2</sub></sub> i<sub>X</sub>ω ∧ i<sub>Y</sub>ω,
</formula>
then
<formula>
i<sub>(W<sub>1</sub> − W<sub>2</sub>)</sub> i<sub>X</sub>ω ∧ i<sub>Y</sub>ω =  0        ∀X,Y
</formula>
is fulfilled. It is possible only when
<formula>
W<sub>1</sub> − W <sub>2</sub> = <under><ope>∑</ope><sat>s</sat></under>v<sub>s</sub> ∧ u<sub>s</sub>
</formula>
where <math>v<sub>s</sub></math> are some vector fields and
<math>i<sub>u<sub>s</sub></sub>ω = 0</math>
(in other words when <math> W<sub>1</sub> − W<sub>2</sub></math> belongs to the class
<math>0<sup>∗</sup></math>)
</paragraph>
<theorem>
 If the non-Hamiltonian vector field <math>E</math>
satisfies <math>[E , X<sub>h</sub><sup>∗</sup>] = 0<sup>∗</sup> </math> commutation
relation (generates non-Noether symmetry), then the functions
<formula>
I <sup>(k)</sup>
= i<sub>W<sup>k</sup></sub>ω<sub>E</sub><sup>k</sup>         k = 1...rank(ω)
</formula>
(where <math>ω<sub> E</sub> = L<sub>E</sub>ω</math>) are constant along trajectories.
</theorem>
<proof>
Let's consider <math> I<sup>(1)</sup></math>
<formula>
L<sub>X<sub>h</sub><sup>∗</sup></sub>I<sup>(1)</sup>
= L<sub>X<sub>h</sub><sup>∗</sup></sub>(i<sub>W</sub>ω<sub>E</sub>)
= i<sub>[X<sub>h</sub><sup>∗</sup> , W]</sub>ω<sub>E</sub> +
i<sub>W</sub>L<sub>X<sub>h</sub><sup>∗</sup></sub>ω<sub>E</sub> = 0
</formula>
The second term vanishes since <math>[E , X<sub>h</sub><sup>∗</sup>] = 0<sup>∗</sup></math> and
<math>L<sub>X<sub>h</sub><sup>∗</sup></sub>ω = 0</math>. The first one is
zero as far as <math> [X<sub>h</sub><sup>∗</sup> , W<sup>∗</sup>] = 0<sup>∗</sup></math> and
<math>[E , 0<sup>∗</sup>] = 0<sup>∗</sup></math> are satisfied. So
<math>I<sup> (1)</sup></math> is conserved.
Similarly one can show that <math>L<sub>X<sub>h</sub></sub>I<sup>(k)</sup> = 0</math> is
fulfilled.
</proof>
<remark>
 <math>W</math> is not unique, but <math>I<sup>(k)</sup></math> doesn't depend
on choosing representative from the class <math>W<sup>∗</sup></math>.
</remark>
<remark>
Theorem is also valid for generators <math>E</math> satisfying
<math> [E , X<sub>h</sub><sup>∗</sup>] = X<sub>f</sub><sup>∗</sup></math>
</remark>
<example>
Hamiltonian description of the relativistic particle leads to the following action
<formula>
A = <ope>∫</ope> p<sub>0</sub>dx<sub>0</sub> + 
<under><ope>∑</ope><sat>s</sat></under>p<sub>s</sub>dx<sub>s</sub>
</formula>
where
<math>p<sub>0</sub> = (p<sup>2</sup> + m<sup>2</sup>)<sup>1/2</sup></math>
with vanishing canonical Hamiltonian and degenerate 2-form defined by
<formula>
p<sub>0</sub>ω = <under><ope>∑</ope><sat>s</sat></under>(p<sub>s</sub>dp<sub>s</sub> ∧ dx<sub>0</sub> + p<sub>0</sub>dp<sub>s</sub> ∧ dx<sub>s</sub>).
</formula>
<math>ω</math> possesses the "null vector field"
<math>i<sub>u</sub>ω = 0</math>
<formula>
u = p<sub>0</sub><fraction><num>∂</num><den>∂x<sub>0</sub></den></fraction> + <under><ope>∑</ope><sat>s</sat></under>p<sub>s</sub><fraction><num>∂</num><den>∂x<sub>s</sub></den></fraction>.
</formula>
One can check that the following non- Hamiltonian vector field
<formula>
E =p<sub>0</sub>x<sub>0</sub><fraction><num>∂</num><den>∂x<sub>0</sub></den></fraction>
+ p<sub>1</sub>x<sub>1</sub><fraction><num>∂</num><den>∂x<sub>1</sub></den></fraction> + ⋯ + p<sub>n</sub>x<sub>n</sub><fraction><num>∂</num><den>∂x<sub>n</sub></den></fraction>
</formula>
generates non-Noether symmetry. Indeed, <math>E</math> satisfies
<math>[E , X<sub>h</sub><sup>∗</sup>] = 0<sup>∗</sup></math> because of
<math>X<sub>h</sub><sup>∗</sup> = 0<sup>∗</sup></math> and <math>[E,u] = u</math>.
Corresponding integrals of motion are combinations of momenta:
<formula>
I<sup>(1)</sup> = <under><ope>∑</ope><sat>s</sat></under>p<sub>s</sub><line />
I<sup>(2)</sup> = <under><ope>∑</ope><sat>r &gt; s</sat></under> p<sub>r</sub>p<sub>s</sub><line />
 ⋯ <line />
I<sup>(n)</sup> = <under><ope>∏</ope><sat>s</sat></under>p<sub>s</sub>
</formula>
This example shows that the set of conserved quantities can be obtained from a single
one-parameter group of non-Noether transformations.
</example>
<acknowledgements>
Author is grateful to Z. Giunashvili and M. Maziashvili for
constructive discussions and particularly grateful to George Jorjadze for invaluable help.
This work was supported by INTAS (00-00561)
and Scholarship from World Federation of Scientists.
</acknowledgements>


<references>
<cite xml:id="r1">
	<who>S. Hojman</who>
	<what>A new conservation law constructed without using either Lagrangians or Hamiltonians</what>
	<where>J. Phys. A: Math. Gen. 25 L291-295</where>
	<when>1992</when>
</cite>
<cite xml:id="r2">
	<who>F. González-Gascón</who>
	<what>Geometric foundations of a new conservation law discovered by Hojman</what>
	<where>J. Phys. A: Math. Gen. 27 L59-60</where>
	<when>1994</when>
</cite>
<cite xml:id="r3">
	<who>M. Lutzky</who>
	<what>Remarks on a recent theorem about conserved quantities</what>
	<where>J. Phys. A: Math. Gen. 28 L637-638</where>
	<when>1995</when>
</cite>
<cite xml:id="r4">
	<who>M. Lutzky</who>
	<what>New derivation of a conserved quantity for Lagrangian systems</what>
	<where>J. Phys. A: Math. Gen. 15 L721-722</where>
	<when>1998</when>
</cite>
<cite xml:id="r5">
	<who>N.M.J. Woodhouse</who>
	<what>Geometric Quantization</what>
	<where>Claredon, Oxford</where>
	<when>1992.</when>
</cite>
<cite xml:id="r6" eprint="math-ph/0106018">
	<who>G. Chavchanidze</who>
	<what>Bi-Hamiltonian structure as a shadow of non-Noether symmetry</what>
	<where>math-ph/0106018</where>
	<when>2001</when>
</cite>
</references>
</article>