The Method of Fuzzy Logic and Data Mining for Monitoring Troposphere Parameters Using Ground-Based Radiometric Complex

The paper presents an algorithm for processing data obtained with the help of ground-based water vapor radiometer (WVR) under conditions of intense precipitation. Fuzzy logic technology and method of singular spectrum analysis were used for realization of this algorithm. Data mining is used to support the choice of the most relevant variables for membership functions of fuzzy logic and inference rules. The selected variables are used as inputs to a fuzzy logic system designed to filter the signal from noise in a quasi-real time scale and predict data during exposure to intensive precipitation. The results of computer modelling of the algorithm demonstrating its good accuracy are presented. The results of comparing integral water vapor content obtained with WVR and Global Navigation Satellite System (GNSS) are discussed.