Digital transformation of machine technology for onion harvesting using the theory of artificial neural networks

• Main Authors: A. V. Sibiriev, A. S. Dorokhov, A. G. Aksenov
• Format: Article
• Language: Russian
• Published: Federal Agricultural Research Center of the North-East named N.V. Rudnitsky 2019-02-01
• Series: Аграрная наука Евро-Северо-Востока
• Subjects: modeling; dynamic system; mathematical model; qualitative indicators
• Online Access: https://www.agronauka-sv.ru/jour/article/view/305

The problem of forecasting the qualitative indicators of onion harvesters was solved using the methodologies of the system analysis and synthesis, physical modeling, based on the theory of artificial neural networks. Analysis of the mathematical model of the working process of onion harvesting machine showed that the increase in the quality indicators of onion harvesting can be ensured by the optimal ratio of internal unregulated parameters of separate executive devices. A change in the process parameters of mechanical means for onion harvesting within design limits does not ensure keeping to agrotechnical requirements. This neural network model for the assessment of quality indicators of functioning elements of the machine for harvesting onion set allows to predict the quality performance indicators on the basis of a large number of external impacts X, affecting the harvesting process. The theory of artificial neural networks allows to describe the technological working process of the machine for harvesting onion set, its individual functioning elements, to predict and evaluate the quality performance indicators both of separate executive devices and the entire machine.