Experimental design for solving inverse heat conductivity problems

A numerical method is proposed for determining the optimum arrangement of a fixed number of temperature sensors when solving inverse heat conductivity problems. The computational algorithm is developed using a spline approximation of the functions to be identified and the formalism of sensitivity functions. The determinant of a normalized Fisher matrix is used as a criterion for selecting the sensor coordinator vector. The coefficients of the matrix are determined by solving the corresponding boundary value problems. The approach proposed here is illustrated by an example.