Singular integral equations

Publication date 6 February 2000

The book represents a well-written and by examples richly illustrated textbook for those who want to be introduced into the topic ``Singular and convolution integral equations. Although the book cannot be used for deepening the knowledge in particular classes of integral equations, it might be very useful for advanced engineers and students. A special feature, distinguishing the book from others written on integral equations, are simple exposition (minimal number of lengthy proofs), ample examples and exercises to each particular class of integral equations. This is especially important for those who would like to work out habits of practical solution of integral equations. The authors supply by examples almost all principal sections and demonstrate practical use of the exposed theoretical material. Bibliographical notes and exercises are concluding each chapter. Let us survey the content of the book. In Chapter 1 the authors expose reference material -- improper integrals, the gamma and the beta functions, the Lebesgue integral, Cauchy principal value and Hadamard finite part integrals, the Fourier and the Laplace transforms. In Chapter 2 ``Abel's and related integral equations the authors present the basic theory of the equations \[ f(s)=\int\limits_a^s\frac{g(t)}{(t-s)^\alpha}ds ,\quad 0<\alpha<1 , \quad a<s<b\tag{1} \] (namely, two different methods of explicit solution) and of some related ones, also when \(a\to-\infty\), \(b\to\infty\) and when Re\((\alpha)<1\). Similar equations are considered on closed and open curves on the complex plane. Chapter 3 ``Cauchy type integral equations the authors start with direct methods of solution to an equation \[ Sg(s):=\int\limits_0^1\frac{g(t)}{t-s}dt=f(s) , \quad 0<s<1 \tag{2} \] where the integral is understood in the Cauchy principal value sense. Afterwards, they present an explicit solution to a singular integral equation with constant coefficients \(g-\lambda Sg=f\), \(\lambda=\)const on the interval \([0,1]\). The same equation, but on a closed smooth contour \(C\), is solved in \S 3.5. In the next two sections the authors discuss analytic representation of functions by Cauchy integral in the domains bounded by the contour \(C\), the Plemelj-Sokhotsky and the Poincaré-Bertrand formulas and they investigate special sectionally-analytic functions \((z-a)^{n-\nu}(z-a)^{m-\mu}\). The material is applied to Cauchy integral equations with constant coefficients on an open arc and on finite union of disjoint open arcs. Special attention is given to the case when the underlying contour extends to infinity. In conclusion of the Chapter, equations with the Hilbert kernel \[ ag(s)+\frac{b}{2\pi}\int\limits_0^{2\pi}g(t)\cot\frac{t-s}2dt=f(s) , \quad 0<s<2\pi\tag{3} \] and constant coefficients are solved. In Chapter 4 the authors discuss singular integral equations with variable Hölder continuous coefficients \[ \alpha(\xi)g(\xi)+\beta(\xi)\int\limits_C\frac{\gamma(\omega) g(\omega)}{\omega-\xi}dt , \quad \xi\in C ,\tag{4} \] when the range of integration \(C\) is either a closed contour or an open arc. The authors call (4) the Carleman equation. In \S 4.2 the authors pick up a special case, \(\beta(\xi)=\lambda=\)const, \(\gamma(\xi)\equiv 1\), while in \S 4.3 they discuss the Riemann-Hilbert problem for the sectionally-analytic function \[ \Phi_1(\xi)Y_+(\xi)-\Phi_2(\xi)Y_-(\xi)=F(\xi) , \quad \xi\in C ,\tag{5} \] which is solved by a standard factorization method. The more general case (\(\gamma(\xi)\equiv 1\)) is considered in \S 4.4. The non-normal case of problem (5), when \(\Phi_1(\xi)\) and \(\Phi_2(\xi)\) might have zeros, is discussed in \S 4.5. The most general Carleman equation is investigated in \S 4.6 by reducing it to the system of simpler Carleman equations. An operational approach to some particular Carleman equations is applied in \S 4.7, while in \S 4.8 the authors consider Bückner's integral equation \[ \int\limits_0^1\left(\frac 1{t-s}+\frac p{t-s}\right)\psi(t)dt=\varphi(s) , \quad 0<s<1\tag{6} \] by relating it to the Cauchy type equation (1) on \([-1,1]\). In Chapter 5 the authors discuss distributional solutions of singular integral equations on the interval \([a,b]\), \(-\infty\leq a<b\leq+\infty\). Two types of test-function spaces are considered: either bounded or vanishing at the endpoints with all derivatives up to \(\infty\). Abel's and related equations are considered in \S\S 5.3-5.4. To operators of fractional integration and to Cauchy integral equations the next two sections are devoted. Analytic representation of distributions, the case of disjoint intervals and equations involving periodic distributions are concluding topics of the chapter. In Chapter 6 the authors continue the investigation of distributional solutions, but now they pick up the entire real line \([-\infty,\infty]\). They treat integral equations with the Hilbert transform \[ a(x)g(x)+\frac{b(x)}\pi\int\limits_{-\infty}^\infty\frac{g(t)}{t-x}dt=f(x) , \quad -\infty<x<\infty.\tag{7} \] The authors identify several subspaces of the space \(S'\) of distributions where the Hilbert transform \(H\) is an isomorphism. Asymptotic estimates of distributional solutions are discussed and non-normal equations are treated. In Chapter 7 integral equations with logarithmic kernels \[ \int\limits_a^b[P(x-y)\ln|x-y|+Q(x,y)]g(y)dy=f(x) , \quad a<x<b,\tag{8} \] with polynomial \(P\) and degenerated kernel \(Q(x,y)\) are solved in ordinary and generalised spaces of functions (starting with the case \(P\equiv 1\), \(Q=0\)). Equations with such kernels arise in boundary value problems for elliptic partial differential equations in 2-dimensional domains (the potential method) and can be reduced to Cauchy singular integral equations by differentiation, but the equivalent transition needs careful treatment. In conclusion, asymptotic behaviour of eigenvalues of (8) with \(P\equiv 1\), \(Q=0\) is discussed. In Chapter 8 the classical Wiener-Hopf equation \[ g(x)+\int\limits_0^\infty k(x-y)g(y)dy=f(x) , \quad 0<x<\infty\tag{9} \] with absolutely integrable kernel \(k\in L_1(-\infty,\infty)\) is solved. The authors apply factorisation and again consider both regular and generalised solutions. In the last Chapter 9, dual integral equations are investigated \[ \int\limits_a^b K_j(x,y)f(y)dy=A_j(x) , \quad x\in{\mathcal I}_j , \quad j=1,2,\tag{10} \] where \({\mathcal I}_1=[a,c]\) and \({\mathcal I}_2=[c,b]\). The equations are reduced to one on the interval \({\mathcal I}_1\) by applying a representation of the solution. By another method the authors find resolvent kernels for \(K_1(x,y)\) and \(K_2(x,y)\) on the corresponding intervals \({\mathcal I}_1\) and \({\mathcal I}_2\). To implement the above-described approaches, the Hankel transform is involved. In \S 9.3--\S 9.5 equations with trigonometric kernels, Beltrami's dual integral equation and triple integral equations are treated by the first approach. The concluding three sections deal with the dual equations of Titchmarsh type by applying the second method in the space of regular functions and also in the space of distributions \(H_\mu'\) defined there.

Recommendations

Cited in

(67)

This page was built for publication: Singular integral equations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4939494)