An approach to constructing linearized input/output mappings of multi-input/multi-output stochastic systems is presented. The approach is based on involving an information-theoretic measure of dependence of random vectors as a criterion… Click to show full abstract
An approach to constructing linearized input/output mappings of multi-input/multi-output stochastic systems is presented. The approach is based on involving an information-theoretic measure of dependence of random vectors as a criterion of selection of input variables of the model. An application of the mathematical technique developed to a task of advanced automated process control system functions regarding nuclear power plants efficiency monitoring is considered.
               
Click one of the above tabs to view related content.