AN ADAPTIVE SYSTEM FOR STABILIZING UNSTABLE CRYPTOCURRENCY RATE BASED ON A COGNITIVE MAP IMPULSE PROCESS MODEL

In this paper a cognitive map (CM) of the cryptocurrency in the financial market has been developed based on causal relationships. This is a weighted directed graph with nodes reflecting: the cryptocurrency rate, the volume of cryptocurrency trading, the number of cryptocurrency users, the volume of capitalization, the volume of investments, the volume of speculation in cryptocurrency, supply and demand for cryptocurrency, indirect profit, level of confidence in cryptocurrency, variance of the cryptocurrency rate, integral level of risks when using cryptocurrency. On the basis of the CM, a dynamic model of CM impulse processes is developed in the form of a system of difference equations (Roberts equations). The external control vector for the CM impulse process is selected. It is formed by varying the resources of the following CM nodes coordinates: cryptocurrency supply, trading, capitalization, investment and speculation volumes. Controls are designed in a closed-loop control system of the impulse process based on the automatic control theory methods and are implemented by a decision-maker. A closed-loop control system for the CM impulse process is implemented, which includes a multidimensional discrete state controller synthesized on the basis of the modal control method, which generates the selected control vector and directly affects respective CM nodes by varying their coordinates. The paper solves the problem of a stabilization system design for the unstable transient process of the cryptocurrency rate based on the modal control method. An identification system for the CM weights based on the recurrent least squares method has been developed. The options of modal control with two, three and five controls in the CM with 12 nodes are investigated.