On the construction of a variational principle for a certain class of differential-difference operator equations (Q6186373): Difference between revisions

From MaRDI portal
Added link to MaRDI item.
ReferenceBot (talk | contribs)
Changed an Item
 
Property / cites work
 
Property / cites work: Q5547044 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5343704 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Variational principles for nonpotential operators / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5601116 / rank
 
Normal rank
Property / cites work
 
Property / cites work: On variational formulations for functional differential equations / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4235108 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q2719450 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Elliptic functional differential equations and applications / rank
 
Normal rank

Latest revision as of 11:21, 26 August 2024

scientific article; zbMATH DE number 7798854
Language Label Description Also known as
English
On the construction of a variational principle for a certain class of differential-difference operator equations
scientific article; zbMATH DE number 7798854

    Statements

    On the construction of a variational principle for a certain class of differential-difference operator equations (English)
    0 references
    0 references
    2 February 2024
    0 references
    In general terms, the operator equation \[ N(u) = 0\quad (u \in D(N)) \] has a \textit{variational formulation} if the solutions (the zeros of \(N(u))\) can be characterized as critical points of a functional \(F_N(u)\) called the \textit{potential} of \(N(u);\) correspondingly, \(N(u)\) is the \textit{gradient} of \(F_N(u).\) In the setting of this paper, \(N: D(N) \subseteq U \to V\) with \(U\) and \(V\) real normed linear spaces. As per standard definitions, the \textit{Gâteaux variation} of \(N\) is \(\delta N(u, h) = \lim_{\epsilon \to 0} (N(u + \epsilon h) - N(u))/ \epsilon\) where the limit exists. If \(\delta N(u, h)\) is a linear operator of \(h\) then it is called the \textit{Gâteaux differential} of \(N.\) The operator \(N\) is called a \textit{potential operator} with respect to the bilinear form \(\Phi(\cdot, \cdot) : V \times U \to\mathbb{R}\) if there exists a functional \(F_N : D(F_N) = D(N) \to\mathbb{R}\) such that \[ \delta F_N(u, h) = \Phi(N(u), h) \quad \big(u \in D(N), h \in D(\delta N(u, \cdot))\big). \] In this case, the problem is said to have a \textit{direct variational formulation,} and the problem of constructing \(F_N\) from \(N\) is called the \textit{classical inverse problem} of calculus of variations. This theoretical setup is applied to the operator generated (in suitable function spaces) by the left hand side of a difference-differential equation of arbitrary degree \(n\) and arbitrary number of time delays \(\omega_1, \dots, \omega_l:\) \[ \begin{multlined} \mathcal{F}(t, u(t), u(t - \omega_1),\dots, u(t - \omega_l), u'(t), u'(t - \omega_1),\dots, u't - \omega_l), \dots, \\ u^{(n)}(t), u^{(n)}(t - \omega_1),\dots, u^{(n)}(t - \omega_l)) = 0 \end{multlined} \] including functional initial conditions. The author gives conditions on the equation that make the results applicable and illustrates this with several examples.
    0 references
    differential-difference equations
    0 references
    calculus of variations
    0 references
    variational principle
    0 references
    potential operator
    0 references
    criterion of potentiality
    0 references
    direct variational formulation
    0 references
    inverse problem
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references