METHOD FOR NORMALIZATION OF INPUT DATA OF AUTOMATED ARTIFICIAL NEURON SYSTEM OF TECHNICAL CONTROL

摘要

Method for normalization of artificial neuron system of technical control includes peripheral devices blocks and central processor with system of linear normalization of input data for teaching artificial neuron systems, peripheral devices block of linear pre-normalization, block of non-linear normalization with variable type of non-linearity, and central processor with system of determination of coefficients and type of non-linear normalization and evaluation of accuracy of results of normalization this automates control of peripheral devices blocks that in sequence perform bringing heterogeneous input data to range set by operator at achieving accuracy parameters set by operator through scanning input data, calculation of parameters pf linear transformation and type of non-linear transformation, according to parameters of neuron network that performs data processing and sequential change of parameters of non-linear normalization of input data to range and distribution with given neuron network and accuracy set by operator.