Technological adjustment of agricultural machines based on fuzzy logic

Introduction. The search for optimal values of the adjustable parameters of a combine harvester in the field is a complex challenge. Both improving the design of the machine and using of automated systems based on fuzzy control increases the quality of harvesting. The article describes information support for the preliminary technological adjustment of complex harvesting machines that operate in changing field conditions. The object of research is a combine harvester. Materials and Methods. We analyzed the quantitative, qualitative and estimated information during the technological adjustment of the harvesting machine. We used a logical-linguistic approach and a mathematical apparatus of fuzzy logic to find the optimal values of the parameters. The composition of fuzzy relationships between the semantic spaces of external factors and the controlled parameters of the machine was used as the basis of the mechanism for the logical derivation of solutions. The developed paradigm of decision-making based on fuzzy expert knowledge includes the stages of fuzzification, composition and defuzzification. MATLAB environment and Fuzzy Logic Toolbox software were used for calculations. Results. The questions of creation of the expert knowledge base, a quantitative evaluation of the consistency of expert information intended for further deductive inference of solutions in various problems of preliminary tuning are considered. The proposed decision-making scheme is illustrated by the example of selecting the values of the rotation frequency of the separator fan. This is one of the most important adjustable parameters. Models of environmental factors and adjustable parameters of the combine are constructed in the form of semantic spaces and their corresponding membership functions. The generalized domain model has the form: R = X - Y, where R is the fuzzy relation "environmental factors - adjustment parameters" R [Xt, T ( X, ), U, G, M } x [Yj, T(Yj), U, G, M} ; V( x, y) e X x Y; X and Yt are linguistic variables; T is plurality of values of the linguistic variable, or terms, which are here fuzzy variables defined on a plurality of U; G is syntactic procedure describing the process of formation of a plurality of T new values of the linguistic variable; M is a semantic procedure that allows each new value generated by procedure G to be displayed in a fuzzy variable. A database of production rules for fuzzy inference is created and its fragment is given for one of the crops. Conclusions. Application of the logical-linguistic approach to solving the problem of preliminary tuning of machines makes it possible to take into account all types of quantitative, qualitative and heuristic information about the external environment. This ensures the maximum adequacy of the description of the actual harvesting conditions and the optimal-ity of the decisions taken on the settings based on expert information.