Processing data on changes in water quality indicators in water supply systems

The purpose of this work is to consider and apply tools for processing heterogeneous data obtained at key measuring points of a centralized water supply network using cloud technologies, to develop an algorithm for processing this data, as well as the results of the algorithm's operation at one of the centralized water supply facilities. Practical solutions for the application of the considered information approaches for monitoring the technical and sanitary condition of a centralized water supply system are presented. A structural scheme of functioning has been developed and one copy of the data collection system on the state of cold and hot water supply has been assembled. This instance was installed in one of the boilers of the city of Murom (Vladimir region, Russian Federation). OWEN equipment was used for automated data collection. Data collection was carried out via 4G mobile communication channels. Data on the condition of pipes and water quality is accessed via the OWEN Cloud cloud service. The integrated moving average (ARIMA) statistical autoregression model was used to analyze and predict time series. For this model, the average absolute percentage error (MAPE), the RMS error (RMSE), taking into account the square of the deviations, and the coefficient of determination (R² ) were calculated.