Barbaruk L. Models and method for large data processing in integrated monitoring system of water bodies

The thesis is devoted to the actual scientific and applied task of increase of efficiency of integrated monitoring system of water bodies by means of the development and practical application of the ad-hock models and methods of information technology for large data processing. The object of research is the processes of large data processing in integrated monitoring system of water bodies, and the subject of research is the models and method of information technology of processing and visualization of large data used in integrated monitoring system. The analysis of the current state and trends in the development of water monitoring systems shows the further need to develop effective models and methods of information technology for large data processing. The main scientific results are: A new model for analytical large data processing of water monitoring system based on formalization of its attributes and interpretation of uncertainty of water quality assessment in the form of linguistic variables was obtained, which allowed to provide integrated characteristics of fisheries water bodies for further decision making. The method of fuzzy clustering for large data has been improved, which, in contrast to the existing ones, generalizes the procedure of automatic labelling of fuzzy clusters obtained using heuristic algorithms for intuitive fuzzy sets, which allowed the use of automatic mark-up in processing large data and improve convergence of the clustering algorithm. The technology of visualization of large data has been improved by means of correlation matrices and the technology of simplification of polygonal chains, which allowed carrying out dynamic visualization and reducing the time for analysis of data obtained from the monitoring system, including long recordings. The information technology for large data processing in integrated monitoring system of water bodies by its adaptation to tasks of control and management of fisheries enterprises, which provides a semantic basis for integrated automation in terms of implementation of basic analytical functions. This enables to reduce the time to make decisions on the quality of water bodies in fisheries and reduce the time to find discrepancies and retrospective analysis of a large datasets.