Inverse problems for Maxwell's equations

MaRDI QIDQ1907535zbMATH OpenFDO

Authors Vladimir G. Romanov, Sergey I. Kabanikhin

Publication date 13 February 1996

Published in Inverse and Ill-Posed Problems Series (Search for Journal in Brave)

coefficients conditionally stability inverse source problems local well-posedness Maxwell's equations numerical analysis second-kind operator Volterra equations uniqueness

Mathematics Subject Classification ID

Research exposition (monographs, survey articles) pertaining to partial differential equations (35-02) PDEs in connection with optics and electromagnetic theory (35Q60) Inverse problems for PDEs (35R30) Research exposition (monographs, survey articles) pertaining to numerical analysis (65-02) Numerical optimization and variational techniques (65K10) Applications to the sciences (65Z05) Electromagnetic theory (general) (78A25) Inverse problems in geophysics (86A22)

Review text

The main interaction forces on the Earth's surface are gravitation and electromagnetism. Especially electromagnetism plays a very important role in physics, chemistry and biology. Further, our eyes have to interpret electromagnetic waves in order to get information about nature. If scientists wish to interpret measuring data one needs a mathematical model and efficient measuring data (sufficiently many). The mathematical models studied in this book are the Maxwell's equations \[ \text{curl } H= \varepsilon {\partial E\over \partial t}+ \sigma E+ j,\;\text{curl } E= - \mu {\partial H\over \partial t}, \] where \(E= (E_1, E_2, E_3)\) and \(H= (H_1, H_2, H_3)\) are vector functions depending on \(x\in \mathbb{R}^3\) and the time variable. The coefficients \(\varepsilon\), \(\mu\), \(\sigma\) are functions of \(x\in \mathbb{R}^3\) and \(j= j(x, t)\). The direct problem for this system has been extensively studied during the past 100 years. Since the Second World War one has begun (C. Müller) to study such inverse problems from a systematic point of view (radar sets). Especially the unknown parameters \(\varepsilon\), \(\sigma\), \(\mu\) are to be determined by measurements on the Earth's surface (geophysics) respectively outside the human body (medicine). In order to solve an inverse problem the first step consists in studying the direct problem both theoretically and numerically. The next step is to study the information content of an inverse problem, i.e., to find out which inner parameters of the system inaccessible to measurement can be determined in a stable and unique manner. The last step is to develop algorithms for the numerical solution of the inverse problem. The book follows these necessary steps in order to solve inverse problems relative to Maxwell's equations. Chapter 1 deals with the Cauchy problem -- a direct problem. In order to get a feeling for inverse problems the authors study so-called one-dimensional inverse problems, i.e., the coefficients to be determined depend only on one variable (Chapter 2). In this case the electromagnetic field is measured at the plane \(x_3= 0\) and the Fourier transform can be applied to get special solutions. It is very difficult to study inverse problems, when the coefficients depend on two or three variables (Chapter 3). In order to get some results one assumes that \(\varepsilon= \varepsilon_0(x)+ \varepsilon_1(x)\), \(\mu= \mu_0(x)+ \mu_1(x)\), \(\sigma= \sigma_0(x)+ \sigma_1(x)\), where \(\varepsilon_0\), \(\mu_0\), and \(\sigma_0\) are known functions, \(\varepsilon_1\), \(\mu_1\), \(\sigma_1\) sufficiently smooth. Such problems are called linearized problems, where \(\varepsilon_1\), \(\sigma_1\), \(\sigma_1\) are to be determined by calculation. Chapter 4 deals with some special inverse source problems (the right-hand side of the differential equation is to be determined), where \(j(x, t)= \widehat {j} (x) e^{i\omega t}\). In general inverse source problems cannot be uniquely solved without additional information, which follows immediately from the definition of the fundamental solution. In Chapter 5 one can find some inverse problems for quasi-stationary Maxwell's equations. There exist only some special results for anisotropic media, i.e., the coefficients depend also on the direction, \(\varepsilon= (\varepsilon_{ij})\), \(\mu= (\mu_{ij})\), \(\sigma= (\sigma_{ij})\). Such problems are studied in Chapter 6. Numerical methods play a very important role in applications. In general one can get most of the solutions only in this way. We notice that most inverse problems depend discontinuously on the measuring data. In this book the authors essentially use the fact that most inverse problems for hyperbolic equations are reduced to second-kind operator Volterra equations. In particular this property enables the authors to prove local well-posedness of the corresponding inverse problems and to obtain estimates of conditionally stability and the uniqueness theorem on the whole. Inverse problems may be considered among the pressing problems of current mathematical research. It may be that 99.9\% of the mathematicians, physicists and medical doctors do not know the basic principles (results) of interpreting sciences (inverse problems). Therefore the publication of books on inverse problems is highly necessary. These results have to be taught at every university and at every university of technology in order to understand the relationship between nature and physics.

Recommendations

Cited in

(53)

This page was built for publication: Inverse problems for Maxwell's equations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1907535)