The linearization method recently highlighted has been successfully applied in designing an optimal or a flexible control system of a bilinear process actually encountered in case of concentration controls in resources processing. The purposes of the study were to eliminate dependency of control performance upon set point values, to compensate inverse responses of the plant and to extend an allowable range of set point values to be changed. As the plant had to be approximated prior to application of the method, the modelling error was to be considered in the following design stage of a linear controller. Furthermore, state variables were estimated on an approximate nonlinear observer obtained by an extended linearization method, as the variables concerned here could not be detected. The results have been shown in the figures which were obtained exclusively by computer simulations.