Título del libro: 10th 2024 International Conference On Control, Decision And Information Technologies, Codit 2024
Título del capítulo: Outlet Pressure Regulation in a High-Viscosity Two-Phase Flow Horizontal Pipeline Using Inverse ANN and PSO
FLOR LIZETH TORRES ORTIZ; JOSE ENRIQUE GUZMAN VAZQUEZ;
Autores externos:
Idioma:
Año de publicación:
2024
Palabras clave:
Closed loop control systems; Petroleum tar; Pressure regulators; Water pipelines; Air flow-rate; High viscosities; Horizontal pipelines; Input layers; Model-based OPC; Neural network and particle swarm optimizations; Neural-networks; Outlet pressures; Spatial points; Two phases flow; Open loop control
Resumen:
This article presents the development of two models based on artificial neural networks (ANN) aimed at estimating the pressure at a position near the output of a horizontal pipeline that transports a two-phase flow of glycerin-air, which is characterized by its high viscosity. These models consist of an input layer that includes variables such as the air flow rate (Qa), the glycerin flow rate (Qg), and the pressure at three spatial points located before the desired pressure prediction point. In the hidden layer of the first model, the nonlinear activation function Logsig is employed, while in the second model, the Tansig function is used. In both cases, a linear function is used in the output layer. In order to determine the air and glycerin flow rates that were injected to the pipeline, ANN models ares inverted by using the particle swarm optimization (PSO) algorithm. From these inverted models, the design of an open-loop control to regulate the output pressure in the horizontal pipeline is presented. This control allows simultaneous manipulation of the Qg and Qa variables, or a single Qg variable, enabling precise regulation. © 2024 IEEE.
Entidades citadas de la UNAM:
ISBN:
9798350373974
Editorial:
Institute of Electrical and Electronics Engineers Inc.
