CSITA AUTOMATION ISSN 2414-9055 © Computer science, information technology, automation. 2018. Volume 4, issue 3 27 UDC 681.5:622.2 https://doi.org/10.31721/2414-9055.2018.4.3.27 ADAPTIVE CONTROL AND IDENTIFICATION OF DRILLING SYSTEM WITH CONSIDERING THE TYPE OF DRILLED ORE MATERIAL Morkun V.S., ScD, Professor, Tron V.V., PhD, Associate Professor, Paranyuk D.I., MS Kryvyi Rih National University Abstract: The object is to investigate methods for forming a model for the system of adaptive control over drilling with the con- trol object identifier. The following methods have been used in the course of the research: analysis of domestic and foreign experience, systematization of available approaches and methods, methods of numerical simulation for synthesis and analysis of mathematical model, methods of mathematical statistics and probability theory for processing the results of experiments, methods of analytical design and computer simulation in the synthesis and analysis of control system, methods of system analysis in the development of control algorithms. The scientific novelty consists in determining optimal parameters – number of membership functions, membership function type, inputs number - for neuro-fuzzy subsystem of drilling control system with considering the type of drilled ore material. The practical significance of the studies is a structure of adaptive control and identification of drilling system with considering the type of drilled ore material including model of geological structure. The results of the research investigate a strategy of the two-level adaptive control of borehole drilling under rapidly changing conditions which implies simultaneous drilling investigation and control. The subsystem of prediction is implemented on the basis of an adaptive neuro-fuzzy system. The applied neuro-fuzzy system realizes the Sugeno fuzzy inference in the form of a five-layer neural network of signal feedforward, the first layer of which contains the terms of input variables (the current signal value and its delayed values). It should be noted that the membership function type did not influence much on the prediction result. While processing and analyzing the current information about the latest characteristics of drilling and while forming the adaptive control it is reasonable to apply neurofuzzy structures with Gaussian membership functions. Key words: drilling automation, neuro-fuzzy model, adaptive control. Introduction. The creation of control for an object with uncertain parameters is an important problem of the automatic control theory. Nonstationary and uncertain parameters of control objects cause the necessity to create regulators with adaptable parameters ensuring the unchanged accuracy and quality of a system. The creation of the adaptive system with an identifier is aimed at forming a model-identifier of a control object on the basis of fuzzy and incomplete information [1]. Materials and methods 1. Analysis of research and publications The quality of automatic control over technological processes on various stages of iron ore mining and processing can be improved by using the latest information about the technological process while controlling it [2-6, 12- 15]. In this case, the information on the technological process development can be obtained both by its direct measurement and by using a mathematical model [2]. As drilling characteristics are random and non-stationary it is reasonable to apply the methods of adaptive control with an identifier of an object model while synthesizing this process control. The research is aimed at investigating the methods of forming a model for the system of adaptive control over drilling with the control object identifier [7]. In general, when forming the adaptive control of drilling rocks one should consider the fact that the control object is under the influence of the following input impacts: driving X * (t), controlling U(t) and disturbing Z(t). The object’s behaviour characterized by the output variables Y(t) depends on a set of unknown parameters ξ with the given set of admissible values Ξ among which one should distinguish physical and mechanical characteristics of rock types. In this case, it is necessary to form the control that would ensure the given indices of drilling quality under all admissible values of the unknown parameters ξ.