Use of neural network technologies on the example of simulation of the spent oil purification process.

The article presents a model of an ACS (automatic control system) for a spent oil purification process at a diesel locomotive engine, which makes it possible to study the possibility of using a neural network in the process of simulation modeling on the example of modeling the spent oil purification process. Determination of intermediate indicators of the system (parallel and perpendicular speed, coordinate and radius of the impurity particle) to simplify the process of further modeling the system. Simulation modeling was carried out using the developed software package "Model of the Spent Oil Purification Process Based on Pressure Control using a Neural Network" [1]. As a result of the choice of the model and the learning algorithm of the neural network, as well as the simulation, intermediate indicators of the system were determined, such as parallel and perpendicular velocities, coordinate and radius of the impurity particle. In the process of modeling, a control action was determined that allows the actuator in the ACS for the spent oil purification process at the diesel locomotive engine to change the cross-sectional area of the centrifuge inlet. When the cross-sectional area of the centrifuge inlet is changed, the amount of the oil to be purified is affected. Maintaining the constant value of the spent oil flow rate will keep the pressure in the centrifuge within the established parameters. [ABSTRACT FROM AUTHOR]