Development of a control system for a rectification column with a neural network adjustment of the flag liquid flow and upper temperature

Abstract

The gas and oil industries are high-tech industries that are based on modern advances in science and technology. This contributed to the development of the control system and the increase in the level of automation of the oil refining process. At a refinery, which is complex production, a process control system plays a key role. There is a need to provide continuous and error-free control, safe and stable operation of the system. One of the perspectives in control systems is neural networks. They can implement any desired process of a non-linear control algorithm with an incomplete, inaccurate description of the control object, and also provide easily implemented adaptation with unstable statistics. This article presents the development of a control system with a neural network adjustment of the reflux fluid flow rate and the top temperature of the column and a comparison with the number of deviations from the norm of the process parameters under the control of the neuroregulator and under the control of the PID controller. Block diagrams of a control system with a neural network controller and a PID controller are compiled. The expediency of using a neural network controller to control the object is substantiated. A mathematical model of the control object was developed taking into account its inherent internal relationships between the parameters of the technological mode and taking into account the influence of external disturbing factors. Processing of the research results was carried out using the software package “MatLab”.