РЕАЛІЗАЦІЯ ГЕНЕТИЧНОГО АЛГОРИТМУ ШЛЯХОМ ЗАСТОСУВАННЯ ПРОДУКЦІЙНИХ ПРАВИЛ

Today, there are a large number of algorithms for solving clearly structured problems, as well as methods necessary to solve problems for which there are no clear definite steps to find the result. Among the latter, the most common methods of artificial intelligence, and especially “soft methods”, the basis of which is a simplified representation of wildlife. One of them is a genetic algorithm that models evolutionary approaches (crossbreeding, selection and mutation). Genetic algorithms are used to solve complex scientific tasks and problems, which are part of a class of evolutionary algorithms, which are based on biological principles of natural selection. But for every single task of mathematical modeling and optimization, genetic algorithms need to be adapted and modified according to the subject area and the task, which requires the use of an additional complex mathematical apparatus. So, its main drawback is the complex mathematical apparatus that describes this algorithm. An approach is proposed that will circumvent this problem. Its basis is also a method related to artificial intelligence - production rules. Thanks to them, you can move away from the use of complex formulas and focus on solving the task. This article discusses the main stages of the implementation of this approach using an example that describes the use of the genetic algorithm when replacing the opinions of experts in the process of forming membership functions for fuzzy sets.