Neural Network Control Systems for Objects of Periodic Action with Non-linear Time Programs

Abstract

The research is devoted to the development and comparison of one-circuit and combined automatic control systems (ACS), as well as combined ACS with logic devices and ACS with a neural network controller for program control of periodic action apparatuses. The effectiveness of the listed systems was evaluated by the accuracy of the implementation of time programs with non-linear distributions in different variants of transition from one section to another.