Chaotic motion of dynamic systems with “one” degree of freedom

It is shown that the regimes with chaotic motion are also inherent in dynamic systems having only one degree of freedom. Such opportunity arises because such systems in the phase space have areas in which the conditions of the uniqueness theorem are broken, in particular, in the presence of singular...

Full description

Saved in:
Bibliographic Details
Published in:Вопросы атомной науки и техники
Date:2012
Main Author: Buts, V.A.
Format: Article
Language:English
Published: Національний науковий центр «Харківський фізико-технічний інститут» НАН України 2012
Subjects:
Online Access:https://nasplib.isofts.kiev.ua/handle/123456789/108231
Tags: Add Tag
No Tags, Be the first to tag this record!
Journal Title:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Cite this:Chaotic motion of dynamic systems with “one” degree of freedom / V.A. Buts // Вопросы атомной науки и техники. — 2012. — № 1. — С. 328-332. — Бібліогр.: 3 назв. — англ.

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id nasplib_isofts_kiev_ua-123456789-108231
record_format dspace
spelling Buts, V.A.
2016-10-31T20:38:05Z
2016-10-31T20:38:05Z
2012
Chaotic motion of dynamic systems with “one” degree of freedom / V.A. Buts // Вопросы атомной науки и техники. — 2012. — № 1. — С. 328-332. — Бібліогр.: 3 назв. — англ.
1562-6016
PACS: 05.45.-a
https://nasplib.isofts.kiev.ua/handle/123456789/108231
It is shown that the regimes with chaotic motion are also inherent in dynamic systems having only one degree of freedom. Such opportunity arises because such systems in the phase space have areas in which the conditions of the uniqueness theorem are broken, in particular, in the presence of singular solutions. Examples of such systems are given.
Показано, что режимы с хаотическим движением присущи также динамическим системам, имеющим всего одну степень свободы. Такая возможность возникает благодаря тому, что такие системы в своем фазовом пространстве имеют области, в которых нарушаются условия теоремы единственности, в частности, при наличии особых решений. Приведены примеры таких систем.
Показано, що режими з хаотичним рухом властиві також динамічним системам, що мають усього один ступінь свободи. Така можливість виникає завдяки тому, що такі системи у своєму фазовому просторі мають області, у яких порушуються умови теореми єдності, зокрема, при наявності особливих рішень. Наведено приклади таких систем.
en
Національний науковий центр «Харківський фізико-технічний інститут» НАН України
Вопросы атомной науки и техники
Section F. Nonlinear Dynamics and Chaos
Chaotic motion of dynamic systems with “one” degree of freedom
Хаотическое движение систем с "одной'' степенью свободы
Хаотичний рух систем з "одним'' ступенем свободи
Article
published earlier
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
title Chaotic motion of dynamic systems with “one” degree of freedom
spellingShingle Chaotic motion of dynamic systems with “one” degree of freedom
Buts, V.A.
Section F. Nonlinear Dynamics and Chaos
title_short Chaotic motion of dynamic systems with “one” degree of freedom
title_full Chaotic motion of dynamic systems with “one” degree of freedom
title_fullStr Chaotic motion of dynamic systems with “one” degree of freedom
title_full_unstemmed Chaotic motion of dynamic systems with “one” degree of freedom
title_sort chaotic motion of dynamic systems with “one” degree of freedom
author Buts, V.A.
author_facet Buts, V.A.
topic Section F. Nonlinear Dynamics and Chaos
topic_facet Section F. Nonlinear Dynamics and Chaos
publishDate 2012
language English
container_title Вопросы атомной науки и техники
publisher Національний науковий центр «Харківський фізико-технічний інститут» НАН України
format Article
title_alt Хаотическое движение систем с "одной'' степенью свободы
Хаотичний рух систем з "одним'' ступенем свободи
description It is shown that the regimes with chaotic motion are also inherent in dynamic systems having only one degree of freedom. Such opportunity arises because such systems in the phase space have areas in which the conditions of the uniqueness theorem are broken, in particular, in the presence of singular solutions. Examples of such systems are given. Показано, что режимы с хаотическим движением присущи также динамическим системам, имеющим всего одну степень свободы. Такая возможность возникает благодаря тому, что такие системы в своем фазовом пространстве имеют области, в которых нарушаются условия теоремы единственности, в частности, при наличии особых решений. Приведены примеры таких систем. Показано, що режими з хаотичним рухом властиві також динамічним системам, що мають усього один ступінь свободи. Така можливість виникає завдяки тому, що такі системи у своєму фазовому просторі мають області, у яких порушуються умови теореми єдності, зокрема, при наявності особливих рішень. Наведено приклади таких систем.
issn 1562-6016
url https://nasplib.isofts.kiev.ua/handle/123456789/108231
citation_txt Chaotic motion of dynamic systems with “one” degree of freedom / V.A. Buts // Вопросы атомной науки и техники. — 2012. — № 1. — С. 328-332. — Бібліогр.: 3 назв. — англ.
work_keys_str_mv AT butsva chaoticmotionofdynamicsystemswithonedegreeoffreedom
AT butsva haotičeskoedviženiesistemsodnoistepenʹûsvobody
AT butsva haotičniiruhsistemzodnimstupenemsvobodi
first_indexed 2025-11-25T01:11:04Z
last_indexed 2025-11-25T01:11:04Z
_version_ 1850503510714482688
fulltext CHAOTIC MOTION OF DYNAMIC SYSTEMS WITH “ONE” DEGREE OF FREEDOM V.A. Buts∗ National Science Center “Kharkov Institute of Physics and Technology”, 61108, Kharkov, Ukraine (Received October 27, 2011) It is shown that the regimes with chaotic motion are also inherent in dynamic systems having only one degree of freedom. Such opportunity arises because such systems in the phase space have areas in which the conditions of the uniqueness theorem are broken, in particular, in the presence of singular solutions. Examples of such systems are given. PACS: 05.45.-a 1. INTRODUCTION Now there is a common agreement that the regimes with chaotic motion in dynamic systems can exist only if they hold the number of degrees of freedom more or equal than 1.5. The main reason for existence of such opinion is the restriction that in the phase plane (system with one degree of freedom) the phase trajectories should not intersect. In other words, there is a requirement that the theorem of unique- ness be held (see, for example, [1]). If we are limited to a class of ordinary differential equation (ODE) whose parameters satisfy this theo- rem, such statement, certainly, is correct. However, it is known that except the general solutions (for which the theorem of uniqueness is held), ODE have singular solutions. In points of the singular solutions the theorem of uniqueness is not held. Therefore, if we shall take into account dynamic systems in which there are singular solutions, the formulated restric- tions on number of degrees of freedom (1.5) can be removed. In this case the modes with chaotic motion can be realized in the systems with one degree of freedom as well. In the present work we give exam- ples of such systems. 2. SYSTEMS WITH SINGULAR SOLUTIONS Below we shall show that in the systems with one de- gree of freedom at the presence of singular solutions the modes with chaos are possible. It is necessary to note that, as soon as the system is situated in the points of singular solutions, it, in general case, “does not know” its further trajectory. The choice of the further trajectory is determined by any external, even arbitrarily small perturbations. As a character- istic example we shall consider the dynamics of the system which is described by the following equations: ẋ0 = x1; ẋ1 = ( x2 1 2x0 ) − 0.5 · x0. (1) The phase portrait of the system (1) is presented in Fig. 1. The integral curves in this case are circles (see below). Also, the centers of the circles are located on the axis x1, and the radiuses of these circles are equal to distance of these centers from the zero point (x0 = 0; x1 = 0). This point is common for all cir- cles. Besides, this point is the singular solution of the system (1) (see below). The system (1) was analyzed numerically. The results are presented in Figs. 1-6. In the second and third figures the characteristic time dependencies are presented. Comparing Fig. 1 with Fig. 2, it is possible to make the conclusion that the representing point, moving to one of circles after passage of a zero point, finds itself on another circle. Moreover, the transition from one circle to another occurs according to a random law. Really, the spec- trum of this dynamics is wide (see Fig. 3), and the correlation function falls down quickly enough (see Fig. 4). Fig. 1. Phase portrait of the system (1) Let’s note that the casual transitions from one cir- cle to another at passage of the zero point depend on the accuracy of computing. The change, e.g., of the step of calculations changes the concrete character of these transitions. However, as a whole, statistically, the dynamics remains the same. ∗E-mail address: vbuts@kipt.kharkov.ua 328 PROBLEMS OF ATOMIC SCIENCE AND TECHNOLOGY, 2012, N 1. Series: Nuclear Physics Investigations (57), p. 328-332. Fig. 2. Evolution in time of the variable x0. One can see the transitions from one circle to another The system (1) is not unique. Below we shall show how the sets of such systems can be constructed. But now let’s consider the dynamics of one more system, probably, more interesting from the point of view of chaotic dynamics: dx0 dt = x0 · x1 + γ · x1 ≡ F1, dx1 dt = x2 1 − x4 0 − γ · x0 ≡ F2. (2) A phase portrait of the system (2) is presented in Fig. 5. Fig. 3. Spectrum of x0 Fig. 4. Correlation function for variable x0 Addition of linear terms (γ · xi) to the system (2) has allowed to change the character of a special zero- point. From the complex quasisaddle this point has turned into a “centre” point. If such updating is not done, the representing points will remain on the way to this special point during an unlimited time. To get rid of this feature this additional term has been intro- duced into the model. The characteristic dependence of dynamics variable on time is presented in Fig. 6. Fig. 5. Phase portrait of the system (2) at γ = 0 Fig. 6. Characteristic view of realization of va- riable x0 From this figure it is already visible that the dy- namics of transitions from one integrated trajectory to another at the passage in a vicinity of the zero point is complex. Moreover, these transitions occur casually. The randomness of them is visible when we are looking at the spectrum and the correlation function (Figs. 7 and 8). In these two figures one can see that the spectrum is wide enough and the cor- relation function falls down while oscillating. Now let us show how the set of systems with one degree of freedom similar to the systems (1) and (2) can be constructed. Let us take an integral curve which is given by the equation ϕ (x0, x1) = 0. Then the system of equations whose integral will be this integral curve can be represented in the fol- lowing form: dx0 dt = F1 (ϕ, x0, x1) − ∂ϕ ∂x1 M (x0, x1) , dx1 dt = F2 (ϕ, x0, x1) + ∂ϕ ∂x0 M (x0, x1) , (3) where Fs(ϕ, x0, x1) are arbitrary functions which have such property: Fs(0, x0, x1) = 0; M (x0, x1) is an arbitrary function. 329 Fig. 7. Spectrum of x0 Fig. 8. Correlation function for variable x0 Fig. 9. Dependence of amplitude on time Using the system (3), it is possible to construct a large variety of dynamic systems which will possess the necessary properties. As an example we shall con- sider a case that the integral curves are a family of circles with the radius R: ϕ = (x0 − R)2 + x2 1 − R2 = 0. (4) The set of integral curves (4) is presented in Fig. 1. By choosing the functions Fs and M it is pos- sible to achieve an elimination of the parameter R from the system (3). Really, we shall choose (for ex- ample) such functions as: F1 = 0; F2 = ϕ ·f (x0, x1); M = −x0 · f (x0, x1). Here f (x0, x1) is an arbitrary function. Substituting these expressions in the sys- tem (3), we shall get a set of systems of equations in which the parameter R is already eliminated: dx0 dt = 2x0 · x1 · f (x0, x1) , dx0 dt = ( x2 1 − x2 0 ) · f (x0, x1) . (5) Choosing function f (x0, x1) as f (x0, x1) = 1/2x0 we shall get the system of equations (1). The dynamics of this system is chaotic. This is visible from Figs. 2-4. It is necessary to notice the following: the integral curves of the system (5) will be a family of circles with a common point (0, 0). Despite this fact, the character of this point and the dynamics of the system will depend essentially on the form of the function f (x0, x1). Let’s show now that the zero point is a singular solution of the system (1). Really, this point belongs to the family of circles (4). The same circles are inte- gral curves of the system (1). These integral curves are convenient to be rewritten as: x2 1/x0 + x0 = R. From these curves it follows that in the vicinity of zero point the Lipchitz conditions for the system (1) are broken. Really, the Lipchitz condition for the system (1) can be written down as ∣∣∣∣ x̃ 2 1 x̃0 − x2 1 x0 ∣∣∣∣ ≤ L (|x̃0 − x0| + |x̃1 − x1|) , (6) where L is a positive constant. In the vicinity of zero the left part of the inequal- ity (6) can be estimated as ∣∣∣R̃ − R ∣∣∣, where R̃ and R are radiuses of two arbitrary circles. Generally, the differences of these radiuses can be of any size. Thus, in a zero point the Lipchitz conditions are not satis- fied, i.e. the conditions of the theorem of uniqueness for the system (1) are broken. Besides taking the partial derivative of the function (4) with respect to R and equating it to zero, we find that the point (x0 = 0; x1 = 0) is a singular solution of the system (1), and also envelope of the integral curves. Let’s consider now the system of equations (2). If γ = 0 then this system has the following family of integral curves: ϕ = x4 0 + x2 1 − Cx2 0 = 0. (7) Here C is a certain constant. The appearance of these integral curves is pre- sented in Fig. 5. Using this formula the right part of the second equation of the system (2) at γ = 0 can be rewritten in the following form: F2 = −2 · x4 0 + Cx2 0. Using this expression, it is easy to see that the left part of the Lipchitz inequality in a vicinity of small value of x0 will be proportional to x2 0 ∣∣∣(C − C̃) ∣∣∣. Tak- ing into account that the constants C and C̃ also can be essentially different, in a vicinity of zero point (but not namely in this point) the Lipchitz conditions will not be satisfied. However such “cutting down” of the system of equations (2) will behave quite regularly. 330 Moreover, very quickly all the initial points will find themselves in the vicinity of the zero point and will remain in its vicinity arbitrarily long time. To re- move such feature of dynamics in the system (2), the linear terms were added. They do not permit the representing points to stop in the zero point. Let’s point out another opportunity of construc- tion of the systems of differential equations with one degree of freedom whose dynamics can show chaotic behavior. Let us have two families of integral curves ϕ1 (x0, x1) = 0 and ϕ2 (x0, x1) = 0. Using the sys- tem of equations (3), it is easy to find a system of differential equations whose solutions are such inte- gral curves: dx0 dt = ∂ϕ2 ∂x1 F1(ϕ1, x0, x1, t)− −∂ϕ1 ∂x1 F2(ϕ2, x0, x1, t), dx1 dt = −∂ϕ2 ∂x0 F1(ϕ1, x0, x1, t)+ + ∂ϕ1 ∂x0 F2(ϕ2, x0, x1, t). (8) Here Fs(ϕs, x0, x1, t) are arbitrary functions pos- sessing the property Fs(0, x0, x1, t) = 0. From the system (8) it is possible to obtain the following equa- tion for ϕs: dϕs dt = [ ∂ϕ2 ∂x1 ∂ϕ1 ∂x0 − ∂ϕ2 ∂x0 ∂ϕ1 ∂x1 ] Fs ≡ Δ · Fs. (9) From (9) it follows that if we want these integral curves to be stable, the fulfillment of the following inequalities is necessary: [ Δ · ∂Fs ∂ϕs ] ϕs=0 < 0. (10) For our purposes (the occurrence of chaotic dy- namics) is of interest the case that the integral curves ϕ1 (x0, x1) = 0 and ϕ2 (x0, x1) = 0 have a common point. However it is obvious, that if integral curves are stable, and the inequality (10) is fulfilled, the common points will be stable too. In this case, it is obvious that the dynamics will be regular. There- fore it is necessary to require that in common points (in points of crossing of integral curves) the stability must be broken. It can be achieved if one requires [ Δ · ∂Fs ∂ϕs ] ϕs=0 { < 0 ≥ 0. (11) The upper inequality in (11) should be fulfilled in all points, where ϕ1 �= ϕ2. The lower inequality should be fulfilled in the crossing points ϕ1 = ϕ2 = 0. 3. SYSTEM WITH BROKEN CONDITIONS OF UNIQUENESS Above we have considered the systems which have singular solutions. In the points of singular solutions the uniqueness theorem is broken. This fact also has resulted in qualitative change of the dynamics of in- vestigated system. Therefore it is clear that generally there is no necessity in availability of singular solu- tions. It is sufficient that in the phase space of the investigated system there was an area in which the uniqueness theorem is broken. Besides it is important that the trajectories of the investigated system find themselves often enough in this area of phase space. As the elementary example, in which such scenario of the development of chaotic dynamics can be realized, we shall consider the following simple system: ẋ = y, ẏ = −x. (12) Let functions x(t), y(t) be complex. Then two variants of substitutions are possible. The first one is x = xR + i · xI , y = yR + i · yI . In this case the dy- namics of real and imaginary parts are independent. In this case new dynamics does not arise. New dynamics arises at introduction of other de- pendent variables, namely x = x0 · exp (i · x2) y = x1 · exp (i · x3) . (13) Here xk(t) k = {0, 1, 2, 3} are real functions. Substituting (13) in (12) we shall get the following system of ODE for finding xk(t): ẋ0 = x1 cos (x3 − x2) , ẋ1 = −x0 cos (x3 − x2) , ẋ2 = x1 x0 sin (x3 − x2) , ẋ3 = x0 x1 sin (x3 − x2) . (14) These four equations can be rewritten as three ones: ẋ0 = x1 cosΦ, ẋ1 = −x0 cosΦ, Φ̇ = ( x0 x1 − x1 x0 ) · sin Φ, (15) where Φ = x3 − x2. From the first two equations of the system (14) or (15) it is visible that these equations have the follow- ing integral: x2 0 + x2 1 = const. (16) Thus, the system of equations (15) has only one degree of freedom. The system (14) was investigated by numerical methods. The characteristics of their dynamics are presented in Figs. 9-11. From these figures it is visible that as soon as the value of am- plitude appears in the vicinity of zero (x0 = 0), the value of the phase (x2) can undergo a jump. Oc- currence of these jumps is called by the fact that in the point x0 = 0 the theorem of uniqueness for the system (5) is not held. In this point arbitrarily small fluctuations can define the value of a phase. As a matter of fact, in the given elementary example the choice is not so rich: the phase can either re- main continuous or undergo a jump to a value of π. The occurrence of these jumps, as it is visible from Fig. 10, is a stochastic function. Similar dynamics is observed for amplitude x1 and for a phase x3. As a result of these random jumps the dynamics of all system becomes chaotic. The degree of random- ness can be characterized by correlation function (see 331 Fig. 11). For our case the correlation function falls down quickly enough. It is possible to show that also other statistical characteristics for the system (14) (spectra, main Lyapunov index) are the same as for the systems with chaotic dynamics. Fig. 10. Evolution of phase in time Fig. 11. Correlation function for variable x0 4. CONCLUSIONS Thus, if we abandon the conditions of uniqueness of the solutions, then in such systems the modes with chaotic motion can appear. The simplest systems in this case are the systems with one degree of freedom which we have considered above. However, it is clear that in the systems with a large number of degrees of freedom such mechanism of occurrence of chaotic regimes will appear too. It is necessary to say that in many systems which, for example, are described by the systems of equations (1) or (12) the random- ness arises as a result of the presence of small fluctua- tions at numerical calculation. And, the more precise calculation is used, the smaller area in a vicinity of zero will define the chaotic motion. Thus, chaotic dynamics of these systems is practically caused by the presence (even arbitrarily small) of casual forces. Just for this reason the word “one” was taken in in- verted commas in the title of this paper. However, these cases, apparently, do not exhaust all opportu- nities. If the dynamics of the system is such as the area of non-uniqueness in the phase space is not lim- ited to one point, the chaotic dynamics of the system may not depend on numerical fluctuation. However, these questions require an additional study. It is nec- essary to notice that some aspects of the dynamics considered above are similar to the dynamics of sin- gular point mapping which were studied, for example, in [2, 3]. References 1. A.J. Liechtenberg, M.A. Lieberman. Regular and Stochastic Motion. New York: “Springer- Verlag New York inc.”, 1983, 499 p. 2. S.V. Slipushenko, A.V. Tur, V.V. Yanovsky. Mechanism of intermittency arising in singular conservative systems // Izvestia vuzov “AND”. 2010, v. 18, N. 4, p. 91-110. 3. S.V. Slipushenko, A.V. Tur, V.V. Yanovsky. Intermittency without chaotic phases // Func- tional materials. 2006, v. 13, N. 4, p. 551-557. �������� �� ���� �� ������ � �� ��� ����� �� ����� � ���� ��� ��������� � � �� �� � ��� ���� � �� ��� �� � ��� ���� � ��� ���� � � � ����� ���� � ����� ���� � ����� �������� ����� ��������� � ���� ��� ������� � ���� � � �� � � � ��� � ���� �� ������ � �� ��� �� ��� ����� � � �� � �� �� �!�� �� ����� � �� ��� �� �� ������ � � ��� ��� � � ��� � ������ �!�� "� � ������ � �� � �� � � � ��� ������ �� ��� ������ � �� ��� ����� �� ����� � ���� ��� ��������� �� �� � � ��� �� � ���� ���� �# ���� � ���#�� � � � ����� �� ��� � ������ �� � � ��#�� ������ � ���� ���� �#� � � � ��$ ������ ���� �� ��# � � �� � ���$�� ������� � �� � # ��� � ����� #� � �� � �� �!�� ��� ���� �� �� $���� #� ��� ���� � ������� # ����� � � #!���� %������� � ���� �� � � � ��� &&'