Method of minimizing the risk of reducing the production of agricultural products by means of fuzzy logic

摘要

The results of the analysis and prediction of climatic changes in the empirical time series of precipitation regimes and air temperature regime of the initial observations of meteorological stations located in the territory of the Kabardino-Balkarian Republic for the period 1955-2017 are presented. Estimates of random errors and statistical significance of distribution parameters, homogeneity (according to the criteria of Dixon and Smirnov-Grubbs) and stability (according to Student's and Fisher's criteria) during the time trends of integral empirical distributions of meteorological characteristics are obtained. Using the values of the yield of selected crops for the 30 years period, a computer adaptive fuzzy-logical model of the dependence of crop yield on changes in natural factors has been developed. By setting the predicted values of the climatic characteristics in the computer model, we obtain the possible values of the yield for the next agricultural year. In case of unsatisfactory yield values, we determine the analogue-year from previous years in accordance with the next year and develop recommendations for making management decisions.