Information technology components selecting for equipment critical applications

The paper is devoted to the problem of the active components (transistors) automated diagnostics on the base of his current-voltage characteristics and passive radio components (resistors) the prediction class by his parameters. The procedure of identification on the neuron nets base for these transistors is proposed. For automated diagnostic of passive radio components, for example, resistors, the procedure of prediction of class such components is work out. On the base of these procedures and classification and clustering on the wavelet transformation base the information technology of components selection is created. This information technology have to increase the efficiency of passive components rejection (at the expense of production test time) and improve the efficiency of grading active components (by reducing the time of diagnosis on a reduced parameter vector) with sufficient reliability for these tasks.