Prediction of safety in emergency situations of complex geotechnical objects using neural networks

Abstract

Listen

Translate article

In this article, the mathematical setting of the problem of predicting the state of an object was performed, the construction features and efficiency of using a multilayer perseptron for constructing neural network models and predicting the states of a geotechnical object were analyzed. Based on this article, an approach to predicting maximum water levels in rivers during the spring flood is proposed. Thus, prevent and minimize the likelihood of emergencies arising from flooding of territories located in the immediate vicinity of rivers.