Minimization of residual function by quasi-gradient method for identification of the hydraulic conductivity in case of three-dimensional anisotropic aquifer

The identification problem of the hydraulic conductivity in the three-dimensional anisotropic confined aquifer is solved in case of one-phase filtration according to the head values in observation locations. A new quasi-gradient algorithm has been proposed for minimization of a residual function. The sensitivity storage distribution is used for constructing this algorithm. The properties of a residual function in minimum point have been analyzed. These properties are the basis of a new stopping criterion. In this criterion the knowledge of errors in head data is not required.