MULTI-STAGE OPTIMIZATION OF A COMPLEX TERRITORIAL DISTRIBUTED SYSTEMS

Abstract

One of the main trends in the development of information technology and information systems (IT and IS) in the 21st century will be the solution to the problem of the comprehensive integration of these technologies and systems with existing and future production and socioeconomic structures and appropriate management systems (SAs). Therefore, one of the urgent scientific problems is the task of optimizing the management of complex geographically distributed systems, which include production and economic systems. The main type of production and economic system is a modern enterprise, that is, an economic entity. Such an entity can be agroholding, an oil and gas company, an energy complex, information systems, sectoral management systems, large banking structures and similar complex geographically distributed systems.  The main aim of the research is the creation of new intellectual information technologies for the optimization of logistic processes in the conditions of agricultural holdings , improve the accuracy of forecasting of the main economic indicators of economic entities, improvement of investment attractiveness of national agricultural holdings by reducing costs in the logistics process, reducing the risks in investment activities. Applied evolutionary methods, methods and algorithms of project management, fuzzy sets, and neural network algorithms of group account of data handling  (GMDH). The analysis of the usefulness of the logistics activities of the agricultural holding as a project. Identify the means of improving the efficiency of solving this problem. Developed new technologies for optimization of logistics processes through the application of two-stage methods. Tested the performance of the proposed intelligent technologies on the data of economic entities of different degrees of integration. The results indicate that the proposed two-stage optimization technique using the apparatus of fuzzy sets to generate the initial populations of the genetic algorithms gives a more qualitative and quantitative result. The use of neural network method of group account of arguments allows to optimize the costs of future periods and reduce storage costs of commodity products. The expediency of representation of the production activity of the agricultural holding as a project. Developed intelligent information technology for optimization of logistics processes based on the application of modified genetic algorithms involving fuzzy sets to generate the initial populations. Based on the proposed intelligent information technologies were carried out calculations of the cost of the logistics process at several national holdings. The results obtained show a decrease in expenditure in the range of 8 to 12 percent, in terms of integration into international structures are essential.

DOI: https://doi.org/10.28925/2414-0325.2017.3.37z87