Журнал "Проблеми економіки" 2024, № 3

Permanent URI for this collection

https://dr.csbc.edu.ua/handle/123456789/932

Browse

Browsing Журнал "Проблеми економіки" 2024, № 3 by Author "Войніков, Микола Юрійович"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Supply Chain Management Using Evolutionary Algorithms
    (Науково-дослідний центр індустріальних проблем розвитку НАН України, 2024) Скіцько, Володимир Іванович; Войніков, Микола Юрійович
    The era of digital transformation has made it possible to accumulate large amounts of data that can be used in the decision-making process, in particular in supply chain management. With the complication of the problems to be solved, classical optimization methods lose their effectiveness and do not allow to obtain a solution in an acceptable time, which creates the need to study another suitable tools, among which there are evolutionary algorithms that use the principles of biological evolution, allowing to obtain solutions close to optimal (or even exactly optimal) in an acceptable time. Evolutionary algorithms are part of a broader field in artificial intelligence that is evolutionary computing. The article allocates the characteristics of evolutionary algorithms that distinguish them from other algorithms of evolutionary computing, and analyzes the most popular evolutionary algorithms: genetic algorithm, genetic programming, evolutionary programming, evolutionary strategies and differential evolution, in particular, their features and areas of application in supply chain management. A comparative analysis is carried out and recommendations are provided for the selection of the appropriate algorithm, taking into account the characteristics of the problem, in particular, the structure of the solution (coding), the discreteness or continuity of variables, and the speed of getting into the local optimum. The available literature is analyzed and a list of the use of various evolutionary algorithms for the tasks of supply chain management is provided, in particular, in warehouse planning, transportation organization, work planning, etc. Since the effectiveness of the application of evolutionary algorithms depends not only on the choice of a specific algorithm, but also on the choice of parameters, their flexible configuration, etc., in future studies it is advisable to consider modifications of evolutionary algorithms, both hybrid and adaptive approaches.