Utilization of genetic algorithms in problems of discrete optimization of corroding structures

Abstract

The paper proposes a new efficient algorithm to solve the problems of corroding hinged-rod structures optimal design, which involves obtaining solutions with given accuracy. When solving the optimization problem, it is required to determine the parameters of cross-sections of elements in such way that the volume of the structure is minimal, and for a given period of operation it retains its carrying capacity, that is, it satisfies the constraints on strength and stability. Calculation of the constraint functions involves a numerical solution of the system of differential equations describing the process of corrosion in structural elements. The effect of mechanical stresses on the rate of corrosion involves solving the problem of the stress-strain state at each node of the time grid, which leads to increased requirements for the efficiency of the computational algorithm. Hingedrod structures are made of rolling profiles; the dimensions of sections of these profiles are regulated by standards. Therefore the search for solution is made in a discrete non-metric space of varied parameters. The optimization algorithm is based on use of flexible tolerance strategy together with integer-valued real genetic algorithm. Use of the flexible tolerance method allows to change the accuracy of a solution at different stages when solving the optimization problem using the genetic algorithm. To ensure the required accuracy of the restrictions function computation in the vicinity of an extremum an artificial neural network approximating the relationship between structure parameters, permissible error of solution (flexible tolerance criterion) and parameters of computational procedures is used. The analysis of effectiveness of the developed algorithm, which used the number of calls to the procedure of the finite element method in the process of searching for the optimal solution as the efficiency criterion, demonstrated a significant decrease in computational costs in comparison with the known algorithms.

Authors and Affiliations

О. Р. Денисюк, С. А. Борзов

Keywords

Related Articles

Математическое моделирование сборки составной оболочки

Сборка является заключительным этапом изготовления машин, аккумулирующим накопленные на предшествующих этапах несовершенства. Для составной оболоч- ки многоступенчатых ракет контейнерного базирования существующий уровень...

ПІДХОДИ КЛАСТЕРНОГО АНАЛІЗУ ДО ОЦІНКИ ФІНАНСОВО-ЕКОНОМІЧНОЇ ДІЯЛЬНОСТІ ПІДПРИЄМСТВ

Стаття присвячена фінансово-економічному оціннюванні стану автотранспортних підприємств України шляхом використання багатовимірного статистичного кластерного аналізу. В статті було розв’язано задачу аналізу фінансово-еко...

Calculation algorithm of operational chacteristics of adsorptive heat storage device for decentralized space heating system

The work is devoted to the development of an efficient algorithm for solving the problem of calculating the operating characteristics of an adsorption heat energy storage device for a decentralized space heating system....

DEVELOPMENT OF AN ANALYTICAL MODEL FOR CALCULATING AND OPTIMIZING VIBROACOUSTIC LOADING OF TECHNOLOGICAL EQUIPMENT PIPELINES

The article is devoted to the development of analytical model for calculation and optimization of vibroacoustic load of pipelines of technological apparatuses. A deviation of pipeline geometry from the ideal one and the...

Гнучка виробничо-логістична система: модель управління складом з дефіцитом

Сучасні умови функціонування підприємств в рамках ринкової економіки, а також в умовах невизначеності та нестійкості зовнішнього середовища вимагають нових, високо ефективних засобів та методів управління його господарсь...

Download PDF file
  • EP ID EP642516
  • DOI -
  • Views 114
  • Downloads 0

How To Cite

О. Р. Денисюк, С. А. Борзов (2017). Utilization of genetic algorithms in problems of discrete optimization of corroding structures. Комп’ютерне моделювання: аналіз, управління, оптимізація, 2(2), 14-22. https://europub.co.uk./articles/-A-642516