Model for forecasting yields under fuzzy initial conditions

Abstract

The problems of constructing a forecasting model for yields under indistinctly specified information on the climatic and agrotechnical conditions of growing agricultural crops have been considered. A yield forecasting model based on fuzzy models of the Sugeno type has been described. The model and rules of fuzzy inference have been presented in an unclear knowledge base in the form of an expert matrix of knowledge. An algorithm for constructing a prediction model of this type has been given. The organization of a computational experiment has been described, according to the effectiveness of the proposed model for forecasting the yield of cotton.