Dynamic chaos generated by linear systems

It is shown that regimes with chaotic behavior are also inherent in linear systems. A dynamic chaos which can arise in essentially quantum systems (not quasi-classical) is of special interest. In particular, it is shown that the diffusion of quantum systems in the energy space can be considerably mo...

Full description

Saved in:
Bibliographic Details
Date:2012
Main Author: Buts, V.A.
Format: Article
Language:English
Published: Національний науковий центр «Харківський фізико-технічний інститут» НАН України 2012
Series:Вопросы атомной науки и техники
Subjects:
Online Access:http://dspace.nbuv.gov.ua/handle/123456789/108230
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:Dynamic chaos generated by linear systems / V.A. Buts // Вопросы атомной науки и техники. — 2012. — № 1. — С. 333-336. — Бібліогр.: 7 назв. — англ.

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id irk-123456789-108230
record_format dspace
spelling irk-123456789-1082302016-11-01T03:02:41Z Dynamic chaos generated by linear systems Buts, V.A. Section F. Nonlinear Dynamics and Chaos It is shown that regimes with chaotic behavior are also inherent in linear systems. A dynamic chaos which can arise in essentially quantum systems (not quasi-classical) is of special interest. In particular, it is shown that the diffusion of quantum systems in the energy space can be considerably more effective than transitions at multiphoton processes. It is shown that taking into account the singular solutions allows to realize regimes with chaotic behavior even in systems with "one'' degree of freedom. Показано, что режимы с хаотическим поведением присущи также линейным динамическим системам. Особый интерес представляет динамический хаос, который может возникать в существенно квантовых системах (не квазиклассических). В частности, показано, что диффузия квантовых систем в пространстве энергии может быть значительно более эффективной, чем переходы при многофотонных процессах. Показано, что учет особых решений позволяет реализовать режимы с хаотическим поведением даже в системах с «одной» степенью свободы. Показано, що режими з хаотичним поводженням властиві також лінійним динамічним системам. Особливий інтерес представляє динамічний хаос, що може виникати в істотно квантових системах (не квазікласичних). Зокрема, показано, що дифузія квантових систем у просторі енергії може бути значно більш ефективною, ніж переходи при багатофотонних процесах. Показано, що облік особливих рішень дозволяє реалізувати режими з хаотичним поводженням навіть у системах з «одним» ступенем свободи. 2012 Article Dynamic chaos generated by linear systems / V.A. Buts // Вопросы атомной науки и техники. — 2012. — № 1. — С. 333-336. — Бібліогр.: 7 назв. — англ. 1562-6016 PACS: 05.45.-a; 05.45.Mt http://dspace.nbuv.gov.ua/handle/123456789/108230 en Вопросы атомной науки и техники Національний науковий центр «Харківський фізико-технічний інститут» НАН України
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
language English
topic Section F. Nonlinear Dynamics and Chaos
Section F. Nonlinear Dynamics and Chaos
spellingShingle Section F. Nonlinear Dynamics and Chaos
Section F. Nonlinear Dynamics and Chaos
Buts, V.A.
Dynamic chaos generated by linear systems
Вопросы атомной науки и техники
description It is shown that regimes with chaotic behavior are also inherent in linear systems. A dynamic chaos which can arise in essentially quantum systems (not quasi-classical) is of special interest. In particular, it is shown that the diffusion of quantum systems in the energy space can be considerably more effective than transitions at multiphoton processes. It is shown that taking into account the singular solutions allows to realize regimes with chaotic behavior even in systems with "one'' degree of freedom.
format Article
author Buts, V.A.
author_facet Buts, V.A.
author_sort Buts, V.A.
title Dynamic chaos generated by linear systems
title_short Dynamic chaos generated by linear systems
title_full Dynamic chaos generated by linear systems
title_fullStr Dynamic chaos generated by linear systems
title_full_unstemmed Dynamic chaos generated by linear systems
title_sort dynamic chaos generated by linear systems
publisher Національний науковий центр «Харківський фізико-технічний інститут» НАН України
publishDate 2012
topic_facet Section F. Nonlinear Dynamics and Chaos
url http://dspace.nbuv.gov.ua/handle/123456789/108230
citation_txt Dynamic chaos generated by linear systems / V.A. Buts // Вопросы атомной науки и техники. — 2012. — № 1. — С. 333-336. — Бібліогр.: 7 назв. — англ.
series Вопросы атомной науки и техники
work_keys_str_mv AT butsva dynamicchaosgeneratedbylinearsystems
first_indexed 2025-07-07T21:10:32Z
last_indexed 2025-07-07T21:10:32Z
_version_ 1837024027682537472
fulltext DYNAMIC CHAOS GENERATED BY LINEAR SYSTEMS V.A. Buts∗ National Science Center “Kharkov Institute of Physics and Technology”, 61108, Kharkov, Ukraine (Received October 27, 2011) It is shown that regimes with chaotic behavior are also inherent in linear systems. A dynamic chaos which can arise in essentially quantum systems (not quasi-classical) is of special interest. In particular, it is shown that the diffusion of quantum systems in the energy space can be considerably more effective than transitions at multiphoton processes. It is shown that taking into account the singular solutions allows to realize regimes with chaotic behavior even in systems with “one” degree of freedom. PACS: 05.45.-a; 05.45.Mt 1. INTRODUCTION Practically till now it is considered that the regimes with dynamic chaos (DC) arise only in nonlinear sys- tems. For this reason nobody carries out the search of regimes with DC at the study of linear systems. However, it is well known that by the transformation of the dependent variable the linear systems often can be transformed into nonlinear systems. In such systems regimes with DC are widely presented. Well- known examples are transitions from the equations of quantum mechanics to the equations of classical me- chanics, and also transitions from the Maxwell equa- tions to the equations of geometrical optics. Thus, today there are known at least two exam- ples at which in linear systems at certain values of their parameters the regimes with DC are possible. In works [1-3] it is shown that this situation is con- siderably more widespread and that the regimes with dynamic chaos are inherent to a large number of lin- ear systems. In the present work the results of the analysis of this feature for a number of classical and quantum systems are given from unified positions. A possible mechanism of occurrence of unpredictability is discussed. For these purposes it is convenient to use the concept of a measure. The introduction of stochastic conjugated functions is also useful. These functions are irregular. However, some of their com- binations behave regularly. The examples of such functions are given. The significant role of singular solutions in the origin of chaotic dynamics is shown. 2. DYNAMIC CHAOS AT INTERACTION OF THREE LINEAR OSCILLATORS For definiteness we shall consider the most simple and at the same time one of the most important linear physical system, in which the mode with DC is possi- ble. This system is three connected linear oscillators: q̈0 + q0 = −μ1q1 − μ2q2; q̈1 + q1 = −μ1q0; q̈2 + (1 + δ) q2 = −μ2q0, (1) where q̇ ≡ dq dτ , μi � 1, δ � 1. The r.h.s terms (factors of connection) of the sys- tem are small. That is why the solutions of the sys- tem (1) are convenient to be searched as: qi = Ai(τ) exp (iωit) . (2) For finding complex amplitudes Ai(τ) it is possi- ble to get the following system of equations: 2iȦ0 = −μ1A1 − μ2A2 exp(iδτ), 2iȦ1 = −μ1A0, 2iȦ2 = −μ2A0 exp(−iδτ). (3) The system of equations (3) is linear. For further analysis of dynamics of Ai(τ) we shall present them as: Ai(τ) = ai(τ) exp(iϕ(τ)), (4) here ai, ϕi are real amplitudes and real phases. Sub- stituting (4) in (3) for finding ai and ϕi we shall get the following equations: ȧ0 = − (μ1/2)a1 · sin Φ − (μ2/2) · a2 · sin Φ1; ȧ1 = (μ1/2)a0 · sin Φ; ȧ2 = (μ2/2)a0 · sin Φ1; Φ̇ = (μ1 2 )( a0 a1 − a1 a0 ) cosΦ − (μ2 2 )( a2 a0 ) cosΦ1, Φ̇1 = (μ2 2 )( a0 a2 − a2 a0 ) cosΦ1 − (μ1 2 )( a1 a0 ) cosΦ+δ, (5) where Φ ≡ ϕ1 − ϕ0, Φ1 ≡ ϕ2 − ϕ0 + δτ . The system (5) is already nonlinear. Basically, the dynamics of such system can be chaotic. The analysis of system (5) gives the following estimation for the condition of occurrence of regimes with DC: (μ1 + μ2) > δ. It is possible to find more detailed information about the system (5), and also about the results of numerical investigation in [1-5]. Let us note ∗E-mail address: vbuts@kipt.kharkov.ua PROBLEMS OF ATOMIC SCIENCE AND TECHNOLOGY, 2012, N 1. Series: Nuclear Physics Investigations (57), p. 333-336. 333 only that at fulfilling the conditions for onset of DC the dynamics of amplitudes (ai) and phases (ϕi) had been found as chaotic. The spectra are wide, correla- tion functions fall down quickly, the main Lyapunov indexes are positive. 3. WAVES SCATTERING IN THE MEDIUM WITH WEAK PERIODIC INHOMOGENEITY Properties similar to the considered case of regular or chaotic dynamics in the system of three linear cou- pled oscillators take place in a large number of other linear systems. Here we consider a model of an elec- tromagnetic wave scattering on a non-uniform dielec- tric medium filling the half-space (z > 0 for unam- biguity). The wave comes from the upper (z < 0) homogeneous half-space. The wave of electrical field satisfies the well known wave equation: ΔE + k2εE = 0, where k = ω/c. We consider the case that the permittivity is de- scribed by the formula: ε = 1 + 2∑ i=1 μi · cos (�κi · �r) . It is assumed that the heterogeneity is small, μi � 1. In this case the electromagnetic wave scat- tering gives diffracted waves of the minus first order of diffraction. The complete field can be represented in the form: E = 2∑ i=0 Ei = 2∑ i=0 Ai(z) · exp ( i�ki�r ) . Here the first term corresponds to the incident wave, the second and third ones correspond to the minus first order of diffraction by the first and the second heterogeneities accordingly. Let the relation between the wave vectors satisfy the following expres- sions: �k1 = �k0 − �κ1, �k2 = �k0 − �κ2 + �δ, �δ = (0, 0, δ) . In this model we consider the case that all the wave amplitudes depend only on the coordinate z di- rected into the lower half- space. Substituting equa- tion for the field into the wave equation and applying averaging to find the slow varying amplitudes of the interactive waves we obtain the following system of equations: 2iA′ 0 = μ1A1 + μ2A2 exp(iδz), 2iA′ 1 = μ1A0 · (k0z/k1z) , 2iA′ 2 = μ2A0 · (k0z/k2z) exp(−iδz), (6) where A′ ≡ dA/dz and the following dimension- less parameters and independent variables are intro- duced: μ1 ≡ μ1 · k/k0z, μ2 ≡ μ2 · k/k0z, δ ≡ δ/k, z ≡ k · z. The system (6) is similar to the system of equa- tions (3) if the derivative with respect to time is sub- stituted by the derivative with respect to the coor- dinate z. Therefore, the dynamics of the systems (3) and (6) has the same qualitative description. They both have the areas of parameters in which their behaviour is chaotic. The chaos criterion, for example for the system (6), in the symmetric case (μ1 ≈ μ2 = μ), is given by the inequality: δ < (k2 · μ)/4 √ k1z · k2z . 4. CHAOS IN QUANTUM SYSTEMS Anyone who uses words “quantum” and “chaos” in the same sentence should be hung by his thumbs on a tree in the park behind the Niels Bohr Institute Joseph Ford This epigraph (from [6]) reflects the attitude of many scientists to researches on quantum chaos. This atti- tude, basically, is caused by the fact that the equa- tions of quantum mechanics are linear equations. Be- low we shall show that at the appropriate transforma- tion of dependent variable the equations of quantum mechanics become nonlinear, in which the regimes with DC exist. Let us consider a quantum system which is under perturbation. Its Hamiltonian is Ĥ = Ĥ0 + Ĥ1(t). We will use the conventional theory of perturbation. In this case the solution can be represented in such a form: ψ(t) = ∑ n An(t) · ϕn · exp(iωnt). Unknown amplitudes An(t) must be found from the following system of equations: ih̄ · Ȧn = ∑ m Un m(t) · Am, (7) where Un m = ∫ ϕ∗ m Ĥ1(t)ϕn exp[i t (En − Em)/h̄] dq. Let us consider a more simple case when in our system a harmonic perturbation Ĥ1(t) = Û ·exp(iΩt) is present. In this case the matrix elements of inter- action get such forms: Un m = Vn m exp{i · t · [(En − Em)/h̄+ Ω]}, Vn m = ∫ ϕ∗ n · Û · ϕmdq. (8) Let us consider the dynamics of three- level system (|0〉, |1〉, |2〉) (see the figure). Scheme of energy levels 334 We will suppose that the frequency perturbation and eigenvalue of energies of these levels satisfy such re- lations: m = 1, n = 0, h̄Ω = E1 − E0, m = 2, n = 0, h̄(Ω + δ) = E2 − E0, |δ| � Ω. (9) Using these relations one can limit oneself only to three equations in the system (7): ih̄Ȧ0 = V01A1 + V02A2 exp(iδt), ih̄Ȧ1 = V1 0A0, ih̄Ȧ2 = V2 0A0 exp(−iδt). (10) The system (10) at Vi 0 = V0 i, ( i = 1, 2) is equivalent to the system (3). Let us make such replacement of the dependent variable Ai(τ) = ai(τ) exp(iϕ(τ)). Then for a de- finition of variables ai and ϕi we shall get a sys- tem of nonlinear equations which are equivalent to the system (5). A condition of the development of DC in it will be the inequality (μ1 + μ2) > δ (here μi = 2 ·V0 i/h̄ ·Ω). This inequality has a simple phys- ical sense. It means that Rabi frequency (V0, 1/h̄) of transitions between zero and first levels should be more than the distance between the first and the sec- ond levels. Let us now assume that the conditions of realiza- tion of DC are fulfilled. The investigated systems in this case will casually wander on power levels. It is interesting to give an estimation of time for transi- tion of the system from one level to another and to compare this value to the regular transitions. The dimensionless time in the regular case is equal to the ratio of frequency of external perturbation to Rabi frequency: τr ∼ 1/μ = h̄Ω/U01. We can get such estimation for probability to find our system on first energetic level 〈 a2 1 〉 ∼ μ2 · 〈a2 0 〉 · τ in stochastic regime. The squares of amplitudes vary from 0 to 1. From this one can estimate the average transition time between the levels in a sto- chastic regime by such expression: τch ∼ (τr)2 ∼ (h̄ · Ω/U01) 2. Thus, the time of diffusion in energy space at distance ΔE can be estimated by formula: τD ∼ (ΔE/h̄Ω) (h̄Ω/U01) 2. It is interesting to com- pare this time to the time of multiphoton transitions. The time of multiphoton transitions at distance ΔE can be estimated as: τMPh ∼ (h̄Ω/U01) (ΔE/h̄Ω). If the inequality (ΔE/h̄Ω) � 1 holds, the efficiency of transitions at the stochastic motion will be much higher than the one due to the multiphoton transi- tions: (τD,CH/τMPh) � 1. 5. CHANGE OF PROBABILITY AT TRANSFORMATIONS OF DEPENDENT VARIABLES Such question arises: which replacement of depen- dent variables will lead to a system of nonlinear equa- tions which have regimes with DC? Now there is no exhausting answer to this question. However, there are considerations that the measure and density of probability can be useful in this case. Really, let’s introduce the measure Δμ = p (�xi) · Δ�x. Here Δ�x is the volume of the phase space, p (�xi) is the proba- bility to find our system in the point �xi of the phase space. Let’s make the replacement �zk = f(�xi). Thus the new density of probability is connected to the old density of probability by the following formula (see, for example, [7]): g(�z) = ∑ i p (�xi) Δ�xi Δ�z = ∑ i p (�xi) |J | , (11) here J is the Jacobian of transformation from the old variables to the new ones. If initially we have a linear system, then for its density of probability it is possible to choose function: p(�xi) ∼ δ(�xi − �xi(t)). From the for- mula (11) it follows that we can lose determi- nacy if Jacobian tends to infinity. Let us con- sider which Jacobian has our replacement (5). In our case we have Ak = A′ k + iA′′ k = ak exp (iϕk); A′ k = ak cosϕk, A ′′ k = ak sinϕk. The Jacobian of this transformation is |J | = 1/ |ak|. Thus, if as a result of dynamics of investigated system the am- plitude tends to zero (ak → 0), the Jacobian will infinitely grow. This area of the new phase space will be a source of uncertainty. 6. CONCLUSIONS It is unconditional that the most important result of the carried out researches is the fact that the trans- formation of the dependent variable in linear systems can transform them in nonlinear systems which can have regimes with DC. Let us note that such regimes appeared in the systems which were not exposed to coarsening. Above we saw that the dynamics of new depen- dent variable can be chaotic. At the same time the initial dependent variable obeys the linear equa- tions. Therefore their movement should be regular. Really, the numerical investigation shows that de- spite of the fact that functions ai and ϕi behave ir- regularly being separate, their combination such as A′ k = ak cosϕk, A ′′ k = ak sinϕk behaves quite regu- larly. Such combination of stochastic functions can be named as stochastic conjugated functions. But it is turned out that sometimes the combination of functions A′ k = ak cosϕk, A ′′ k = ak sinϕk behaves irregularly. There is a question, to what is it con- nected? The analysis shows that it occurs only in the case when the amplitudes ai pass through zero. We saw that the information about initial dynamics in this case is lost. However, another factor can be included in the game: the one of violation of unique- ness of the solution. Really, at passage of functions through zero, as it is easy to see, the conditions of the theorem of uniqueness are broken. Looking at all these investigations a question can arise: why is it necessary to transform a simple linear system into a complex nonlinear system? (We have 335 accustomed to the return transformations, when non- linear equations are transformed to linear ones. A classical example now is the method LA pair). The answer to this question is such: first of all at these transformations new sides of dynamics of the investi- gated systems are revealed, and secondly, as soon as we have defined the area of parameters of investigated system, in which the regimes with DC are realized, for the analysis of behavior of investigated system in new variable the methods of statistical physics can be used. In many cases it allows us to understand essentially better the behavior of considered system. In particular, above at the consideration of the dy- namics of a quantum system we have seen that as soon as the Rabi frequency became higher than the frequency between two close levels, the DC develops. Moreover, the efficiency of the diffusion process in the energy space can be much greater than at multipho- ton processes. Already this result justifies the use of such transformations. Above we have considered only one type of trans- formations of dependent variable. Such transforma- tion is most widely used in all physical researches. However it is obvious that other transformations can play a similar role too. As example of such new trans- formation one can point on the one which realizes a transformation of an ordinary second-order linear dif- ferential equation into a Ricatti equation. It is also necessary to pay attention to the fact that now there is a common agreement that chaotic dynamics is possible only in the systems with the number of degrees of freedom equal or more than 1.5. This statement is caused by the fact that there were considered the systems for which the conditions of the uniqueness theorem were fulfilled. The systems having singular solutions were dropped out from the investigation. The systems in which the phase space had areas with violation of the uniqueness theorem were dropped out too. If we take into account such systems then we find out that the chaotic behavior will be inherent even in systems with one degree of freedom. Such example we saw above. More exam- ples and more details of this situation are included in the next work. Acknowledgements The author thanks Yu.L. Bolotin, V.F. Kravchenko, A.A. Sanin, K.N. Stepanov, and V.V. Yanovsky for useful remarks and discussions. References 1. V.A. Buts. Chaotic dynamic of linear systems // Electromagnetic waves and electron systems. 2006, v. 11, N 11, p. 65-70. 2. V.A. Buts and A.G. Nerukh. Elements of chao- tic dynamics in linear systems // The Sixth Inter- national Kharkov Symposium on Physics and En- gineering of Microwaves, Millimeter and Submil- limeter Waves and Workshop on Terahertz Tech- nologies. Kharkov, Ukraine, June 25-30, 2007, v. 1, p. 363-365. 3. V.A. Buts, A.G. Nerukh, N.N. Ruzhytska, and D.A. Nerukh. Wave Chaotic Behaviour Gener- ated by Linear Systems // Optical and Quant. Electronics. 2008, v. 40. p. 587-601. 4. V.A. Buts. True quantum chaos // Problems of Atomic Science and Technology. Series “Plasma Physics”. 2008, v. 14, N 5, p. 120-122. 5. V.A. Buts. Onset of dynamical chaos in quan- tum systems // Progress of Contemporary Ra- dioelectronics. 2009, N. 9, p. 81-86. 6. P. Cvitanović, R. Artuso, R. Mainieri, G. Tan- ner, and G. Vattay. Chaos: Classical and Quan- tum. Copenhagen: Niels Bohr Institute, 2009, http://chaosbook.org/version13/chapters/ ChaosBook.pdf 7. G.G. Malinetsky, A.B. Potapov. Nonlinear dy- namic and chaos. Moscow: “Kombook”, 2006, 240 p. ��������� � ���� ����������� ���� ���� ��������� ���� ��� ��������� � � �� �� � ��� ���� � ������� �� � ��� ���� � ������ � ��� ���� � � � ����� ������ � � �� � ��� ����� � ��� ���� � ����� �� � �� ���� ���� �� � � ����� ����� ���� ���� � � ���� ��� ���� ����� ���� ��� ��� ��� � ��������� � � � !!�� � ���� ���� � � �� � � �� ��" � �� #�� $ ���� �� � ���� ����� ����� #!!�� ����� ��� �� ����� � ���$�!� ����� � �%��" ���� ��������� � � ��� ������ �&�� � �������� ��� ���� � �� �� � ��� ���� � ������� �� ���� � � � ���� � '�����( � �����) �������� ��������� ���� � � ����������� ���� ���� ��������� ���� ��� ��������� �� �� � � ��� �� � ����������� ���� �* ���� �*�*�� � � ���*�� � � � ����� ����" � � � *� � �� � ��� ����+ � ���*�� � ����� �� ���� � � �� � *� � �� ���� �� � � � ���� ��� ���" �*���� �� ��� ,�� ���� ��������� �� � !��*� ���� �� � � � �� � � �� � * ��� $*- ���� �� ������ �*��& �!�� ���)� �*� �� ���� � ��$� �!� ��� � � �%����� ��������� �� ���*� ����� � � *" &��� �������+ ���*���� �� � � ��� �� � ����������� ���* � � � � ���� � '��� �( � ������ ������ � ../