Concrete Design via Stochastic Programming

článek v časopise - ostatní, Jost

angličtina

The goal of the paper is to introduce and discuss new interesting ideas related to the promising application area of nonlinear stochastic programming – reliable concrete design. Mathematical programming applications in civil engineering are quite various as, e.g., structural design, water management, reconstruction of transportation networks, and reliability related computations. Initially, deterministic models have been introduced. The common problem is that these models are naturally complex (involve explicit nonlinearities, logical variables, black-box like terms, more optimization criteria, etc.) in comparison with the available computational power that is still considered quite limited. Therefore, the approximating, simplified but solvable models have been utilized and suboptimal solutions were obtained. However, there is an increasing and serious request from practice to reconsider these traditional approaches and deal with involved uncertainties in less straightforward way. Hence, the search for more realistic risk averse and implementable solutions is the most important challenge in reliable concrete design problems. The paper identifies several important problems that practical applications generate mainly in the area of optimum and reliable concrete design. The traditional deterministic approaches are discussed and evaluated specifically within the context of stochastic programming and their disadvantages are shown. The model reformulations and modifications derived from the stochastic programming principles are introduced and related new ideas are marked and listed. As the next step, model approximations lead to the solvable instances of considered problems and test computations end the paper.

structural design, probabilistic design, stochastic programming

VENCLOVSKÝ, J., ŠTĚPÁNEK, P., LANÍKOVÁ, I., KOKRDA, L., HOŠEK, J.

Institute of Information Theory and Automation of the ASCR

Prague, Czech Republic

0023-5954

Kybernetika

Česká republika

16