Spectral influence functions for boundary effects in multidimensional inverse heat conduction problems

A qualitatively new level of the consideration of a priori information on the relationships to be identified is proposed for the solution of multidimensional inverse heat conduction problems. Spectral influence functions are introduced which describe the dependence of a temperature field on the spectral components of a boundary effect rather than on the boundary effect as a whole. The accuracy of the solution is significantly increased since the influence function contains exact information on object geometry, while compressing information on the distributed boundary conditions.