Intelligent control of gas separation during nitric acid production

Abstract

The article considers the possibilities of modeling the characteristics of gas absorption and cooling processes in the production of nitric acid and controlling this process using the radial basis function and neural networks of direct connection. When training a neural network, the input and output results are obtained from the material balance tables and properties of the absorbing column model to train the network with values. The results obtained using neural network models are mainly compared with the results obtained from modeling books. The result obtained shows that relatively simple neural network models can be used to simulate the stationary state of an absorption column. Using modeling, we will see how the neural network type allows the use of modern management methods.