Abstract
<jats:p>The use of evolutionary genetic AI methods and, in particular, genetic algorithms (GA) ensures the construction of sufficiently universal optimization systems with various architectures and macroparameters. A systematic investigation of GA structures, parameters, and performance has made it possible to identify and synthesise key trends in their modification. The effects of GA structure and parameters on the timing and accuracy of fitness function optimization were performed on the OneMax binary chromosome coding test problem. Numerical studies of the solution of the problem of evolutionary genetic optimization have shown the applicability of a fuzzy controller to increase the efficiencyof GA. Numerical experiments have demonstrated that the average fitness value increases rapidly during the initial stage of the optimization process, which can be attributed to the high initial crossover probability ( Pc ). After 20-50 epochs, the optimization process reaches a stable regime. The fuzzy adaptation of the parameters of the genetic operators makes the algorithm more robust compared to fixed parameters. Recommendations have been formulated for the selection of macroparameters and for substantiating the choice of algorithm modification strategies in specific application domains, including the allocation of limited water resources.</jats:p>