Adaptation of the neural network to the study of electrical characteristics of dispersed materials in sorption processes

Abstract

Abstract Modern technological processes cannot exist without optimizing their parameters. The current direction remains their improvement and identification of such regimes, in which without losing the quality of finished products there is a reduction of energy expenditures and resource saving. Since some processes have a complex structure, and several parameters have to be varied to optimize them, it is advisable to use mathematical devices and models at the primary stage. The subject matter of artificial neural networks refers to the interdisciplinary field of knowledge related to biocibernetics, electronics, applied mathematics, statistics and automation. Artificial neural networks are versatile for developing new technological solutions and analyzing the detected data with the most effective result. The conclusions obtained from the experiments make it possible to understand how the main parameters of sorption processes affect its intensity, as well as to determine the intervals of change of input factors for statistical analysis. All this makes it possible to ensure stability of technological parameters during all experiments and high reproducibility of experiments. This technology is gaining popularity at present due to the growth of computing capabilities and modernization of equipment. Therefore, the direction of research and improvement of all technological parameters and modes is relevant.