Development of an artificial neural network for a combined model of the uranium extraction process
摘要
Based on a conducted literature review, a training sample was compiled. The selected optimal parameters for training an artificial neural network included its structure, activation function, output layer transfer function, and the number of neurons in hidden layers. The results of calculations using the developed artificial neural network have an uncertainty of less than 1%, which confirms its suitability for creating a digital twin of a technological process.