METHODS OF NONLINEAR DYNAMICS IN THE STUDY OF ECONOMIC PROCESSES (ON THE EXAMPLE OF GRAIN PRODUCTION)

This article is devoted to obtaining pre–forecast information based on calculated values of Lyapunov exponents and constructing a long-term forecast of economic indicators of grain production based on the method of artificial intelligence — linear cellular automaton. The following qualitative pre-forecast characteristics of the time series of grain yields in the Stavropol Territory were obtained: the length of longterm memory; a long-term forecast in the form of a numerical expression and in the form of a linguistic variable. Besides, a three-color forecast model was constructed. The fundamental distinction of this method of forecasting time series is the determination of the length of memory. The latter allows you to build a long-term forecast within the found memory length. As practice shows, the magnitude of the prediction error during the validation procedure of a linear cellular automaton does not exceed the threshold of 25 %. Long-term forecasting of economic indicators of grain production in the zone of risky farming is an interdisciplinary problem that which is actively addressed by agronomists, breeders, agrometeorologists, engineers, mathematicians, economists and specialists in other fields. An accurate forecast of the yield of various grain crops and meteorological factors for the next year, as well as the construction of scenarios for the dynamics of grain production development will allow the management to control and regulate the ambitious plans of Russian producers, prescribed in the Long-term strategy [1] for the development of the grain complex of the Russian Federation until 2035. The obtained quantitative values and qualitative pre-forecast characteristics are necessary for the values of the lower level of modeling of grain production activities, which in turn are input information for predictive models of the upper level of grain complex management, for example, when planning the structure of acreage, when hedging grain crops, when determining pricing policy in the external and internal grain market. All of the above becomes a particularly important factor for the development of the domestic grain market in the current sanctions environment.

long-term forecasting, grain production, Lyapunov exponent, linear cellular automaton, sliding control method, memory depth, forecast error, grain yield, cycle, time series, validation, pre-forecast analysis

Kumratova A. M., Sivkov K. A. Methods of nonlinear dynamics in the study of economic processes (on the example of grain production). Business. Education. Law, 2023, no. 1, pp. 72—77. DOI: 10.25683/VOLBI.2023.62.526.