CREATION OF THE RESIT INFORMATION SYSTEM FOR PREDICTION OF RIVER POLLUTION UNDER EMERGENCY SITUATIONS

This article describes developed RESit information system, that created, as improvement to RESit software. System parts are described, such as database, server, interaction between server and possible water flow measurement sensors, learning utility, administration application and user application. RESit advantages are given in relation to using similar systems possibility in emergencies. A practical implementation of such developed methods is described: • pollution forecasting results adjusting method based on a neural network working based on the regression problem; • pollution level forecasting method between specific points using interpolation and recursion methods; • pollution source determining method based on a filtering and sorting algorithm that works on the basis of facts database; • system information with sensors measuring current water flow interaction mechanism. Thus, an information system acquires information technology for forecasting river pollution in emergencies features.