A block model development for intelligent control of the switches operating apparatus position in the electrical interlocking system

Abstract

In this scientific article, we developed a model of block C that controls the position of the switches. A mathematical model of the block was calculated using a two-layer neural network along with the blocks of the block. An imitation model of the switch control block using the SoDeSys program is presented. In the modeling process, it was studied that a multilayer neural network consists of one or more hidden layers of neurons-the entrance, exit, and the neurons located between them. And block C was determined that it was possible to model with the help of a 2-layer neural network and was expressed in the form of the following layers: 1-in the hidden layer: in the C block, the PK and MK relays are listed. This process indicated that the data coming from the PS or PST block would receive the PK and MK releases. At this layer, it is determined in which position the switches are located, and the data is transferred to the next layers. 2-in the hidden layer: the VZ relay in the C Block is indicated. Information coming from PK and MK relays shows the VZ relays acceptance process. In this layer, it is determined whether the switch is cut or not, and information is transferred to the next layers. In this layer, a certain position of the pushed conductor is determined