Management of water flow using long short-term memory algorithm

In this paper, we will overview the program to predict and manage water flow based on the Long short-term memory deep learning algorithm. This article studies water flow management in rivers and canals based on the Long short-term memory algorithm. In the research, an architecture for water flow management was createdf using the Long short-term memory algorithm, the architecture consists of three layers, each layer is justified in the article with a mathematical model, and software was created based on the architecture. An experiment was conducted using the program on the Zarafshan River in the Samarkand region of the Republic of Uzbekistan. A training dataset for predicting water flow was created based on data from the Ministry of Water Resources, and this dataset was used in the experiment. The experiment was conducted over a month, and the parameters time, water flow rate, water pressure, water level, valve coefficient, and water temperature were used to predict and manage water flow. In the study, an experiment was conducted using this software, and the results of the experiment are discussed.