APPLICATION OF A NEURAL NETWORK AND GENETIC ALGORITHM IN THE DESIGN OF MONOLITHIC SLABS ON A BASE

Abstract

The article illustrates the possibility of convergence of mechanics, neurotechnology, and biosimilar technologies. Shown the possibility of using so-called soft-computing in design-related tasks. The work presents the results of self-stresses in the slab on ground obtained using a neural network combined into a system with a genetic algorithm. The possibility of optimizing the geometric parameters of the slab at the given or variable input parameters (strength, self-stress, etc.) by turning on/off artificial genetic features is considered. It has been shown that to describe the state of the structure, where the kinetics of the formation of the concrete structure obeys nonlinear behaviors, the use of neurotechnologies and genetic algorithms is most justified. The article describes the process of developing a neural network and a genetic algorithm, discusses the quality of the solutions obtained.