D-optimal experimental design applied to a linear magnetostatic inverse problem

Abstract

This paper presents the use of D-optimal experimental design to improve the accuracy of the solutions of inverse problems in magnetostatics. This method is applied to the following inverse problem: finding the current density on the basis of magnetic field measurements. The basic idea is to choose the measurement point positions to build a D-optimal experiment, then to solve the inverse problem by iterative regularization. The algorithms used are discussed. Numerical results that show the large improvement in the accuracy of the solutions are presented.