Study on the Simulation Control of Neural Network Algorithm in Thermally Coupled Distillation

Abstract

Thermally coupled distillation is a new energy-saving method, but the traditional thermally coupled distillation simulation calculation process is complicated, and the optimization method based on the traditional simulation process is difficult to obtain a good feasible solution. The neural network algorithm has the advantages of fast learning and can approach nonlinear functions arbitrarily. For the problems in complex process control systems, neural network control does not require cumbersome control structures or precise mathematical models. When training the network, only the input and output samples it needs are given, so that the dynamics of the system can be controlled. In this way, the dynamic performance of the system can be approximated. This method can effectively solve the mathematical model of the thermally coupled distillation process, and quickly obtain the solution of the optimized variables and the objective function. This article summarizes the research progress of artificial neural network and the optimization control of thermally coupled distillation and the application of neural network in thermally coupled distillation.

Application of Artificial Neural Network in Rectification

“…Because of its inherent feedback structure, dynamic recurrent neural network can better express a complex dynamic system and approach the dynamic process of the system with only a single-layer network. Because of its unique structure and information processing method, artificial neural network has obvious advantages in many aspects and has been widely used in various fields, such as robot navigation, natural earthquake prediction [19], intelligent control of high-speed communication network, soil surface pressure measurement [20], rotary deformation verification code recognition, chemical equipment fault diagnosis [21,22], process control and optimization [23][24][25], image recognition [26,27].…”

Application of Artificial Neural Network in Optimal Design of Reactor

“…According to the difference of their functions, artificial neural networks can be divided into feedback network and feedforward network [6]. Feedback neural network generally includes input layer, hidden layer, undertaking layer and output layer [11] [16,17], automatic control [18,19], market analysis [20,21], chemical industry [22][23][24], game theory [25], medicine diagnosis [26][27][28], signal processing [29][30][31], troubleshooting [32,33], machine Learning [34][35][36] and other fields. The construction of artificial neural networks is realized by the simulation of human brain function, rather than by complex mathematical models.…”

A Review of Artificial Neural Networks for Chemical Process Optimization and Compound Property Prediction

“…So far, artificial neural network has been widely used in various fields, mainly including intelligent driving [18,19], auxiliary medical treatment [20,21], voice information processing [22,23], Chinese medicine processing [24,25], chemical process optimization and process control [26][27][28].…”