Purpose. Currently, in the process of solving a wide range of problems in various subject areas, methods for assessing the state of complex objects (SLO) are actively used, among which expert knowledge (EK) occupies a special place. The article proposes a methodological approach to the use of explicit and implicit EK in assessing and predicting the state of the SL. Materials and methods. The article proposes an original fuzzy-feasibility approach to extracting, representing and formalizing the EK in the form of analytical models for assessing the state of the SL. Results. The article describes the essence of solving problems of extraction, formalization, processing and use of explicit and implicit expert knowledge for assessing the state of an SLO at a meaningful level. An example of solving the problem of predicting the level of damage to structural elements and equipment of launch complexes under conditions of intense thermal loading is given. Conclusions. The proposed methodological approach to formalization of explicit and implicit EZ in the form of fuzzy-possibility models has great prospects, since relies on fundamental and applied results obtained so far in theory and practice of application in various subject areas.

The attractiveness of using expert knowledge is obvious, since it uses the experience of highly qualified specialists in this area of knowledge, does not require additional time and money to conduct expensive experiments. However, this approach requires the use of special methods of convolution of expert knowledge (EK) in the multifactor space for solving practical problems. An example of constructing a fuzzypossibility model in a fivedimensional factor space for evaluating a generalized indicator of the efficiency of a technological process for the production of forages from grass is considered. The factors selected by the expert, in turn, are aggregated, i.e., generalized by factors of a lower level of the hierarchy, and constitute a virtual space. Nevertheless, this approach provides an opportunity from a single position and in a single factor space to carry out comparative assessments of the technologies under study. It is shown that the results of quantitative assessment according to the generalized indicator of the effectiveness of technologies completely coincide with the true states of forage production from grasses in farms, depending on the technologies they use – from highintensity to basic ones. The adequacy of calculations for the proposed fuzzypossibility model (NVM) to the actual states of various applied technologies allows us to recommend the considered approach when assessing the state of forage production from grasses for use in agriculture in the NorthWest region of the Russian Federation.

Reindeer is part of the humandeerpasture sociobiological system, where all the constituent parts form emergence. The differences between the reindeer husbandry of foreign countries and Russia are shown. Reindeer in various countries are considered mainly as wild, few herds, and in Russia, at present, 75% of the reindeer population in the Arctic zone belongs to the domestic form. Therefore, the approaches to the study of the problems of reindeer husbandry also differ: the studies of foreign scientists are focused on the study and construction of models of local problems, while Russian studies, in addition to important local problems, are aimed at solving the problems of analyzing the features of the entry of traditional fishing farms into the system of market relations and forms adaptation of traditions to the modern image of economic reality. However, the description of sociobiosystems refers to poorly formalized problems that do not lend themselves to modeling by deterministic mathematical methods, and even more so by statistical ones. Therefore, a new fuzzypossibility approach for constructing multifactor models is proposed, based on the use of explicit and implicit knowledge and experience of experts. An example of constructing an adequate fuzzypossibility model of the stability of a reindeer husbandry in a sevenfactor space of both quantitative and nonquantitative (qualitative) variables is given. The adequacy of the model was confirmed situationally for the state of three reindeerbreeding commercial farms in the YamalNenets Autonomous Okrug for 2019 and 2020.