Factorization of Boundary Value Problems Using the Invariant Embedding Method (eBook)

Factorization Method for Boundary Value Problems by Invariant Embedding presents a new theory for linear elliptic boundary value problems. The authors provide a transformation of the problem in two initial value problems that are uncoupled, enabling you to solve these successively. This method appears similar to the Gauss block factorization of the matrix, obtained in finite dimension after discretization of the problem. This proposed method is comparable to the computation of optimal feedbacks for linear quadratic control problems. - Develops the invariant embedding technique for boundary value problems - Makes a link between control theory, boundary value problems and the Gauss factorization - Presents a new theory for successively solving linear elliptic boundary value problems - Includes a transformation in two initial value problems that are uncoupled

Jacques Henry is Director of Research, emeritus at INRIA Bordeaux Sud-ouest, France. He graduated from Eïcole Polytechnique, Paris (1970). He has worked within INRIA (National Institute for Computer Sciences and Automatic Control, France) since 1974. His work covers control of systems governed by partial differential equations, modeling, parameter estimation and continuation-bifurcation methods applied to biological systems mainly in cardiac electrophysiology and biological sequences comparison. His current interests are on numerical analysis, inverse problems and singular perturbations for partial differential equations. He is developing research on the method of factorization of linear elliptic boundary value problems in terms of product of Cauchy problems. He was leading the INRIA project team Anubis on structured population dynamics. He has a special interest on the evolution of activity of popula- tions of neurons.

---

Factorization Method for Boundary Value Problems by Invariant Embedding presents a new theory for linear elliptic boundary value problems. The authors provide a transformation of the problem in two initial value problems that are uncoupled, enabling you to solve these successively. This method appears similar to the Gauss block factorization of the matrix, obtained in finite dimension after discretization of the problem. This proposed method is comparable to the computation of optimal feedbacks for linear quadratic control problems. - Develops the invariant embedding technique for boundary value problems- Makes a link between control theory, boundary value problems and the Gauss factorization- Presents a new theory for successively solving linear elliptic boundary value problems- Includes a transformation in two initial value problems that are uncoupled