Purpose. To improve the efficiency of mining equipment functioning through the introduction of fuzzy inference algorithms into the control of mining facilities being complex objects with dynamic, randomly varied operation modes. methodology. A stage to determine nonfuzzy input variable systems as statistic indices has been added to univer sally accepted algorithm of a fuzzy interference. The indices make it possible to identify operation mode of mining equipment or operation mode transitions. Each input nonfuzzy variable of the system is a result of statistic processing of information signals from the sensors to find unique regularities within the signal; the regularities should correspond to one or several operation modes of the machine, or transitions between them. Additional linguistic input variables of "Previously, their operation mode was…" are introduced to trace temporally varied operation modes of the mining machine. Taking into consideration features of input nonfuzzy variables formation, the system performs fuzzy infer ence discretely in time; in this context, fuzzy inference interference algorithm is added by the conditions for data ac cumulation, variations, and zero variations in the machine operation mode. Findings. A technique to integrate a behavioural model of a mining machine, considered as a control object, into a fuzzy inference algorithm has been proposed. Moreover, recommendations, concerning determination of both non fuzzy and linguistic input variables of the system, formation of a basis of fuzzy products, and determination of condi tions for data accumulation have been formulated. Originality. Originality of the proposed technique is to use a behavioural model of a mining machine, considered as a control object, for the fuzzy inference algorithm in the form of a set of operation modes of a machine and in the form of a scheme of timedependent changes in operation modes. Practical value. The proposed technique is theoretical basis to solve topical scientific and application problem as well as to use nonfuzzy algorithms to control mining machines to improve their efficiency.