Generalized kinetic equations (Q788887)

The existence and uniqueness theory is developed for the differential equation (1) T f'(x)\(=-A f(x) (0<x<\infty)\) with boundary conditions (2) \(Q_+f(0)=f_+\), \(\lim \sup_{x\to \infty}\| f(x)\|<\infty\) on an abstract Hilbert space H. Here T is a bounded injective self-adjoint Fredholm operator and \(Q_+\) the orthogonal onto the maximal positive T- invariant subspace. The work is a generalization of a work of \textit{R. Beals} [J. Funct. Anal. 34, 1-20 (1979; Zbl 0425.34067)] for bounded A. If A has a zero null space Ker A, one may follow the functional formulation due to \textit{R. J. Hangelbroeck} [Transp. Theory Stat. Phys. 5, 1-85 (1976; Zbl 0363.45010)] to write \(Q_-=I-Q_+\), to introduce the spectral projections \(P_+\) and \(P_-\) of \(T^{-1}A\) for the positive and negative parts of its (real) spectrum and to investigate the invertibility of the operator (3) \(V=Q_+P_++Q_-P_-\), after which the unique solution of boundary value problem (1)-(2) reads (4) \(f(x)=e^{-x^{T-1}A}Ef_+ (0<x<\infty)\), where \(E=V^{-1}\) is the so-called albedo operator. For bounded A, Beals proved the invertibility of V as an operator from \(H_ S\) into \(H_ T\), where \(H_ S\) and \(H_ T\) are completions with respect to suitable inner products. As he also proved their equivalence, he showed Eqs. (1)-(2) to be well-posed in \(H_ T\). For unbounded A the equivalence proof breaks down. In this article it is shown that for all \(f_+\in Q_+[H_ T]\) and injective A there exists a unique solution \(f:[0,\infty)\to H_ S\) of Equations (1)-(2) which satisfies \(f(0)\in H_ T\cap H_ S\), thereby making obsolete the above equivalence statement. A study is made of the relationship between equivalence and boundedness of V. For non-injective A the problem is reduced to a problem with injective A. Following \textit{C. G. Lekkerkerker} [Proc. Edinb. Math. Soc., Sect. A 75 (1975-1976), 283- 295 (1976; Zbl 0335.47034)] \(T^{-1}A\) is first reduced to a suitable complement of the zero root subspace \(Z_ 0=Ker (T^{-1}A)^ 2\). Then, following \textit{C. V. M. van der Mee} [Semigroup and Factorization Methods in Transport Theory, Math. Center Tract. 146 (1981; Zbl 0481.45008)], the operator A is modified on \(Z_ 0\) to get a strictly positive \(A_{\beta}\) and the unique solvability of the corresponding boundary value problem of the type (1)-(2) is exploited to solve the original Eqs. (1)-(2). It then appears that Equations (1)-(2) are always solvable with measure of non-uniqueness being the dimension of a maximal strictly negative subspace of Ker A with respect to the indefinite inner product \([h,k]=(Th,k)\). Further, the relationship is pointed out between the expression (4) and the Case-Van Kampen method of eigenfunction expansion prevailing in neutron transport theory. A pletora of applications is offered to neutron transport, radiative transfer, phonon transport and Sturm-Liouville type diffusion equations.