РАЗРАБОТКА СИСТЕМЫ АВТОМАТИЗИРОВАННОГО СОСТАВЛЕНИЯ ПРОГНОЗА ЛАВИННОЙ ОПАСНОСТИ, С ПРИМЕНЕНИЕМ СОВРЕМЕННЫХ ИКТ И АППАРАТА НЕЙРОКОГНИТИВНЫХ АРХИТЕКТУР

The development of infocommunication technologies opens up great opportunities for the creation of avalanche hazard forecasting systems, the rapid development of containerization and virtualization technologies make it possible to fully use the concept of microservice architecture. The development of the machine learning apparatus opens up wide opportunities for its application to solve the forecasting problem. The use of proven methods can give good results and increase the accuracy of avalanche hazard forecasts. This article discusses: the problem of using modern information and communication technologies and machine learning in the field of avalanche forecasting, data sources for the functioning of an automated avalanche hazard forecasting system. The most important data for use in the operation of such a system are highlighted. There is also a list of requirements that such a system must meet. A prototype of a system for predicting the background level of avalanche danger for the mountain region of Khibiny has been developed. The prototype works using machine learning algorithms. A set of predicted meteorological parameters is used for the operation of the models, the forecast is made for the next day and updated every hour. In addition, an analytical model was launched to predict avalanches from wind slab. At the end of the work, examples of displaying data obtained as a result of machine learning models are given, an example of a message sending information by a Telegram bot is given. Separately, it is necessary to highlight the possibility of integrating the system with existing data collection and analysis systems, the introduction of already proven systems will expand their functionality and provide new opportunities for avalanche forecasters.