[PDF] Renormalized Oscillation Theory for Symplectic Eigenvalue





Previous PDF Next PDF



TRANSITION MATRIX THEORY

12 Nov 2014 DS] 12 Nov 2014 ... MM-Singular transition matrix Ts defined via ? has another feature as follows by the next theorem. Theorem 5.4.



LS.6 Solution Matrices

An equation for fundamental matrices. We have been saying “a” rather than “the” fundamental matrix since the system (1) doesn't have a unique fundamental matrix 



Renormalized Oscillation Theory for Symplectic Eigenvalue

15 Mar 2020 DS] 15 Mar 2020 ... We suppose that the symplectic coefficient matrix of the system ... Sk(?) of system (TS?) is symplectic and.



VMware SD-WAN Edge Platform Specifications

15 Dec 2020 VMware SD-WAN Edge software is zero touch provisioned for a secure optimized ... Edge platform and software release matrix. Software.



mp { IlKf - gl12 + PllLf II21 fK(t-s)f(s)ds=g(t)

and L are upper triangular n x n Toeplitz matrices. Such problems arise when a Volterra convolution equation of the first kind. /. fK(t-s)f(s)ds=g(t)



TS Fuzzy Robust Sampled-Data Control for Nonlinear Systems with

08 Dec 2021 matrix is introduced. Specifically we formulate a reliable controller with state feedback



The Sensitivity of the Matrix Exponential

In this paper we examine how the matrix exponential eAt is affected by pertu. A. Elementary techniques using log v (A t) lie A(t-s)ll lie As 1 ds IA i.



Novel approach to power system state estimation for transmission

In this approach both TS and DS parts of the network can be treated by the same methodology variables; R: covariance matrix of the measurement errors.



The Rigidity Theorems of Hamada and Ohmori Revisited T. S.

of incidence matrices of Hadamard designs apply to a wider class of matrices plement Ds ou the pointnhyperplane design where s = log 2 (v+1). Remark.



Hadamard type fractional time-delay semilinear differential

05 Mar 2020 We propose a delayed Mittag-Leffler type matrix function with logarithm which is an extension ... a (ln ts) n??+1 y (s) ds.



[PDF] Matrices et suites - Lycée dAdultes

19 juil 2021 · Exprimer la matrice ligne B correspondant à la moyenne trimestrielle de maths des élèves e1 e2 e3 à l'aide d'une matrice coefficient C et 



[PDF] Exercices Corrigés Matrices Exercice 1 – Considérons les matrices

Exercice 12 – Soit A et B deux matrices carrées de même ordre on suppose que la matrice AB est inversible d'inverse la matrice C Montrer alors que B est 



[PDF] DS(Matrices)pdf

Terminale S Spécialité Maths Devoir de mathématiques Matrices 16/02/15 – Calculatrices autorisées Observations : NOTE : Exercice 1 :(Sur le sujet)



[PDF] MATRICES - maths et tiques

La somme de A et B est la matrice notée A + B dont les coefficients sont obtenus en additionnant deux à deux des coefficients qui ont la même position dans A 



[PDF] Terminale générale - Matrices et graphes - Exercices - Devoirs

Matrices et graphes – Exercices – Devoirs Terminale Générale - Mathématiques Expertes - Année scolaire 2022/2023 http s ://physique-et-maths fr 



[PDF] Matrices - Exo7 - Cours de mathématiques

La matrice (de taille n × p) dont tous les coefficients sont des zéros est appelée la matrice nulle et est notée 0np ou plus simplement 0 Dans le calcul 



[PDF] Terminale S Spé Maths Titre Cours : Matrices Matrices carrées et

Une matrice est un tableau de p lignes et q colonnes dont les coefficients sont des réels (voir des complexes dans les années futures)



[PDF] MATRICES EXERCICES CORRIGES - Maurimath

MATRICES EXERCICES CORRIGES Exercice n°1 On considère la matrice 2) Donner la valeur de chacun des éléments 14



[PDF] Applications linéaires matrices déterminants

Calculer le produit des matrices ( 1) et ( 2) où 1 et 2 sont deux réels quelconques 2 Montrer que ( ) est inversible et déterminer



[PDF] Feuille dexercices no 6 - Matrices

Feuille d'exercices no 6 - Matrices 1 Calcul matriciel produit de matrices puissances de matrices carrées Exercice 1 (pour s'exercer à la maison) (Voir 

:
Renormalized Oscillation Theory for Symplectic Eigenvalue arXiv:2003.06855v1 [math.DS] 15 Mar 2020 Renormalized Oscillation Theory for Symplectic Eigenvalue Problems with

Nonlinear Dependence on the Spectral Parameter

Julia Elyseeva

Department of Applied Mathematics, Moscow State University of Technology, Vadkovskii per. 3a, 101472,

Moscow, Russia

ARTICLE HISTORY

Compiled March 17, 2020

ABSTRACT

In this paper we establish new renormalized oscillation theorems for discrete symplectic eigenvalue problems with Dirichlet boundary conditions. These theorems present the number of finite eigen- values of the problem in arbitrary interval (a,b] using number of focal points of a transformed conjoined basis associated with Wronskian of two principalsolutions of the symplectic system eval- uated at the endpointsaandb.We suppose that the symplectic coefficient matrix of the system depends nonlinearly on the spectral parameter and that it satisfies certain natural monotonicity assumptions. In our treatment we admit possible oscillations in the coefficients of the symplectic system by incorporating their nonconstant rank with respect to the spectral parameter.

KEYWORDS

Discrete eigenvalue problem; Symplectic difference system; Renormalized oscillation theory;

Comparative index

AMS CLASSIFICATION

39A12; 39A21

1. Introduction

In this paper we consider the discrete symplectic system y k+1(λ) =Sk(λ)yk(λ), k?[0,N]Z, λ?R,(Sλ) with the Dirichlet boundary conditions x

0(λ) = 0 =xN+1(λ),(E0)

where we use the notation [M,N]Z:= [M,N]∩Zfor the discrete interval with endpointsM,N?Z.

The coefficient matrixSk(λ)?R2n×2nof system (Sλ) withn×nblocksA(λ),B(λ),C(λ),D(λ)

depending nonlinearly on the spectral parameterλ?Ris assumed to be symplectic, i.e., for all k?[0,N]Zandλ?Rwe have S

Tk(λ)JSk(λ) =J,Sk(λ) :=?Ak(λ)Bk(λ)

C k(λ)Dk(λ)? ,J:=?0I -I0? .(1.1)

CONTACT Julia Elyseeva. Email: elyseeva@gmail.com

In addition, we assume that the matrixSk(λ) piecewise continuously differentiable inλ?R, i.e., it

is continuous onRand the derivativeSk(λ) :=d dλSk(λ) is piecewise continuous in the parameter λ?Rfor allk?[0,N]Z. Given the above symplectic matrixSk(λ) we consider the monotonicity assumption

Ψ(Sk(λ)) = Ψk(λ) :=JTSk(λ)S-1

k(λ)≥0, k?[0,N]Z, λ?R(1.2) for the 2n×2nmatrix Ψk(λ), which is symmetric for anyk?[0,N]Zandλ?Raccording to [31]. The notationA≥0 means that the matrixAis symmetric and nonnegative definite. Symplectic difference systems (S λ) cover as special cases many important difference equations,such as the second order (or even order) Sturm-Liouville difference equations, symmetric three-term recurrence

equations, and linear Hamiltonian difference systems, see [1, 4, 5, 28, 31]. A complete review of the

history and development of qualitative theory of (S λ) is given in the new monograph [12] (see also the references therein). Classical oscillation theorems connect the oscillation and spectral theories of (Sλ). Assume that we need to know how many eigenvalues of (S

λ), (E0) are located in the given inter-

val (a,b]?R.Then, according to the global oscillations theorems in [6, 8] the difference l d(Y[0](b),0,N+1)-ld(Y[0](a),0,N+1) of the numbers of focal points calculated for theprincipal solutionsY[0](b),Y[0](a) of (Sλ) evaluated at the endpointsλ=aandλ=bpresents the number coefficient matrixSk(λ) and Ψk(λ) in the form S k(λ) =?I0 -λWkI? S k,Ψk(λ)≡Ψk:=?Wk0 0 0? ≥0, k?[0,N]Z, λ?R,(1.3) whereSkis a constant symplectic matrix. The same result for (Sλ) with the general nonlinear

dependence onλwas originally proved in [31] forBk(λ) = const (hereBk(λ) is the block ofSk(λ)

given by (1.1)) and then generalized in [28] to the case rankBk(λ) = const, λ?R, k?[0,N]Z.(1.4) Then it was shown in [18] that assumption (1.4) plays a crucial role in the oscillation theory, in

particular, if (1.4) is violated the number of focal points of the principal solution of (Sλ) loses the

monotonicity with respect toλand then the differenceld(Y[0](b),0,N+ 1)-ld(Y[0](a),0,N+ 1)

can be negative. In this case it is necessary to incorporate oscillations of the blockBk(λ) to present

a proper generalization of the results in [6, 8, 28, 31]. Moreover, it was proven in [18, Corollary 2.5]

that condition holds for someλ0?Rif and only if the real spectrum of (Sλ),(E0) is bounded from below. Observe that for a fixedk?[0,N]Zthe symplectic matrixSk(λ) can be viewed as the fundamental matrix of the linear Hamiltonian differential system (with respect toλ)

Sk(λ) =JΨk(λ)Sk(λ), λ?R,(1.6)

with the symmetric Hamiltonian Ψ k(λ)≥0 given by (1.2). In this context one can introduce the numbers ?(Sk(λ0)) =?k(λ0) := rankBk(λ-0)-rankBk(λ0), k?[0,N]Z(1.7) 2 describing the multiplicities ofproper focal points(see [27]) ofSk(λ)(0I)Tas a conjoined basis of (1.6) and then condition (1.5) means that system (1.6) isnonoscillatoryforλnear-∞. In the recent paper [21] we generalized the results in [18] to the case of symplectic eigenvalue problems with general self-adjoint boundary conditions admitting possible oscillations of their coefficients with respect toλ?R. Therenormalizedand the more generalrelativeoscillation theory makes it possible to replace the differenceld(ˆY(b),0,N+ 1)-ld(Y(a),0,N+ 1) of the numbers of focal points calculated

forλ=aandλ=bby the number of focal points of only one transformed conjoined basis˜Yk(a,b) associated with the WronskianˆYTk(b)JYk(a) ofYk(a) andˆYk(b).Remark that we refer

to the renormalized oscillation theory of (S λ) when the consideration concerns oscillations of the

Wronskian of two conjoined bases of (S

λ) considered for different values ofλ.Therelativeoscillation theory investigates the oscillatory behavior of Wronskians of conjoined bases of two symplectic

systems with different coefficient matricesSk(λ) andˆSk(λ),then all results of the renormalized

theory follow from the relative oscillation theorems for the caseSk(λ) =ˆSk(λ), λ?R, k?[0,N]Z.

The relative oscillation theory was developed for eigenvalue problems for the second order Sturm-Liouville difference and differential equations (with linear dependence onλ) in [2, 3, 23,

29, 33]. In the recent papers [24, 25] the renormalized oscillation theory in [23] is extended to the

case of general linear Hamiltonian systems with block matrix coefficients, which are continuous counterparts of (S The relative oscillation theory for two symplectic problems with Dirichlet boundary conditions under restriction (1.3) is presented in [14, 15], in [16, Theorem 5] the renormalized oscillation theorem for (S λ), (1.3) is extended to the case of general self-adjoint boundary conditions. For the

case of general nonlinear dependence onλthe first results of the relative oscillation theory for two

matrix Sturm-Liouville equations were proved in [17]. In [12, Section 6.1] we presented the relative oscillation theory for two symplectic eigenvalue problemswith nonlinear dependence onλand with general self-adjoint boundary conditions. All these results are derived under restriction (1.4) for S k(λ). The main results of this paper are devoted to the renormalized oscillation theory for (Sλ),(E0) without condition (1.4). In this situation the classical oscillation theorem (see [19]) presents the oscillations ofBk(λ) in terms of numbers (1.7) (see Theorem 2.4 in Section 2) l d(Y[0](b),0,N+ 1)-ld(Y[0](a),0,N+ 1) +? k=0? A similar formula can be proved for the so-calledbackward focal pointsl?d(Y[N+1](λ),0,N+ 1) of the principal solutions atN+ 1 (see Theorem 2.5 in Section 2). The main results of the paper (see Theorems 3.8, 3.12) present renormalized versions of Theorems 2.4, 2.5, respectively. In more details, introducing a fundamental matrixZ[N+1] k(λ) of (Sλ) with the initial conditionZ[N+1]

N+1(λ) =I

we have instead of (1.8) the following renormalized formula l d((Z[N+1](a))-1Y[0](b),0,N+ 1) +? k=0˜ where the numbers ˜?k(ν) are associated with the transformed coefficient matrix

˜Sk(λ) = (Z[N+1]

k+1(a))-1Sk(λ)Z[N+1] k(a) by analogy with (1.7). In (1.9) we have the number l d((Z[N+1](a))-1Y[0](b),0,N+ 1) which describes oscillations of the transformed conjoined basis 3 (Z[N+1] k(a))-1Y[0] k(b) associated with the WronskianY[N+1]T k(a)JY[0] k(b) of the principal solutions Y [N+1] k(a), Y[0] k(b) of (Sλ). The major advantage of using (1.9) instead of (1.8) is the calculation of only one numberld((Z[N+1](a))-1Y[0](b),0,N+1) instead ofld(Y[0](b),0,N+1), ld(Y[0](a),0,N+1)

especially in case of highly oscillatory principal solutionsY[0](b), Y[0](a). The price of this advan-

tage is the necessity to evaluate the second addend in (1.9) which depends on the fundamental matrixZ[N+1](λ) of system (Sλ) evaluated forλ=a.In Section 4 of the paper we decide this

problem presenting (1.9) in an invariant form incorporating oscillations ofSk(a)-Sk(λ), λ?(a,b]

instead of oscillations of blocks of˜Sk(λ).We proved in Section 4 (see Theorem 4.5) that (1.9) is

equivalent to L d((Z[N+1](a))-1Y[0](b),0,N+ 1) +? k=0ρ k(λ) = rank(Sk(a)- Sk(λ-))-rank(Sk(a)- Sk(λ))≥0,(1.10) whereLd((Z[N+1](a))-1Y[0](b),0,N+ 1) is the number of (forward) focal points of a 4n×2n conjoined basis associated withZ[N+1](a))-1Y[0](b) (see Remark 4.6). We call representation (1.10) invariantbecause after the replacement of the matricesSk(λ), k?[0,N]ZbyR-1 k+1Sk(λ)Rkformula (1.10) stays the same. HereRk, k?[0,N+1]Zis an arbitrary sequence of symplectic transformation matrices which do not depend onλ.In the last part of Section 4 we investigate the renormalized oscillation theory for systems (S λ) under the assumptionρk(λ) = 0, k?[0,N]Z, λ?[a,b] which is necessary and sufficient for the equalityLd((Z[N+1](a))-1Y[0](b),0,N+ 1) = #{ν?σ|a < (see Corollary 4.12) under the monotonicity assumptionHk(λ)≥0, k?[0,N]Z, λ?[a,b] for the discrete HamiltonianHk(λ). For the proof of the renormalized theorems we involve new results of the oscillation theory for continuous case - for the differential Hamiltonian systems inform (1.6). Indeed the theory presented

in the paper is now a combination of two oscillation theories- for the discrete and for the continuous

case. Using comparison theorems for the differential case (see [19, Theorem 2.2]) we present a new

interpretation of the results of the discrete spectral theory in [6, 8, 18, 21, 28, 31] which helps to

provide the proof of Theorems 3.8, 3.12 in a compact form. In more details, consider a symplectic

fundamental matrixZk(λ) of (Sλ) associated with the conjoined basisYk(λ) =Zk(λ)(0I)Tunder

the monotonicity assumption Ψ(Zk(λ))≥0 with respect toλ.Then for arbitrary sequence of symplectic matricesRk, k?[0,N+ 1]Zthe matrixR-1 kYk(λ) can be considered as a function of λ?[a,b] with the (continuous) number of proper focal pointslc(R-1 kYk,a,b) for any fixed index kand similarly, as a function of the discrete variablekwith the (discrete) number of focal points l d(R-1Y(λ0),0,N+ 1) for any fixedλ=λ0.Introduce the following closed pathλ=a, k? [0,N+ 1]Z;λ?[a,b], k=N+ 1;λ=b, k?[0,N+ 1]Z;λ?[a,b], k= 0 in the plane (λ,k).Then, according to Theorem 3.5 proved in Section 3 we have the following representation for the sum of all focal points (in the continuous and in the discrete settings) along this path l d(R-1Y(a),0,N+ 1) +lc(R-1

N+1YN+1,a,b)-ld(R-1Y(b),0,N+ 1)-lc(R-10Y0,a,b)

N? k=0l c(˜Sk(0I)T,a,b),(1.11)

wherelc(˜Sk(0I)T,a,b) is the number of focal points of˜Sk(λ)(0I)Tforλ?(a,b] and˜Sk(λ) =

R -1quotesdbs_dbs29.pdfusesText_35
[PDF] ds spé maths

[PDF] sujet bac spé svt glycémie et diabète

[PDF] la lumière des étoiles seconde controle

[PDF] détermination de l élément chimique contenu dans la couche superficielle d une étoile

[PDF] exercice sources lumineuses et spectres

[PDF] controle sur les statistiques 5eme

[PDF] exercice statistique lycée

[PDF] activité statistiques seconde

[PDF] goualard maths

[PDF] exercice graphique 5eme

[PDF] test statistique seconde

[PDF] qu'est ce que le contrôle stratégique

[PDF] le système d'information contribue-t-il ? l'efficacité de la prise de décision

[PDF] maths 1ere es suites exercices corrigés

[PDF] une personne loue une maison ? partir du 1er janvier 1991