Abstract:

The study of mechanics as well as structural analysis in university courses involves the investigation of complex structures and development of their appropriate models. The standard design problem in statics of structures necessarily goes through the stages of numerical and dimension synthesis closely related with the control of strength and stability. A relevant model is usually described algebraically, although this representation is rather simple to obtain only for a structure consisting of a few elements. Undergraduate students learn how to obtain neserssary sets of equations and get analytical solution. Yet the search of an optimal structure meeting the requred limitations becomes much more difficult and almost overwhelming problem in the case of a structure composed from more than ten members. The engineers working in an industry make solutions using highly specified software and even follow their own intuition. The application of genetic algorithms (GAs) is a problem solving strategy that uses stochastic search. They have proven to be particularly useful for solving optimisation problems, thus GAs provide a relatively fast way to simulate complex structures. A considerable advantage for the use of a GAs in university courses related to mechanics and structural analysis is its relatively simple computational implementation along with the dynamically updated visualization snapshots. Thus GAs provide a powerfull teaching means that helps students to make in-depth analysis of relationships between key parameters of a structure and gain knowledge and experience of solving problems of optimization. The example of the use GAs for the optimal design of a coplanar trussed structure is given in the paper.

Keywords:

Genetic algorithm, teaching, mechanics, structure, design, optimization.