Abstract:The mathematical models and solution algorithms for optimization of grillage-type foundations arc presented. Optimization of grillage is based on optimization of separate beams comprising grillage. Minimising of maximum in absolute value vertical reactive force, bending moment, and reaction-bending moment together is sought in a separate beam. All these problems are non-linear, therefore are solved iteratively changing in each iteration the structure shape to a better neighbouring shape. Solution of this requi… Show more
“…The optimal scheme of grillage should possess the minimum possible number of piles. Theoretically, reactive forces in all piles should approach the limit magnitudes of reactions for them [23]. This goal can be achieved by choosing appropriate pile positions.…”
Section: Experimental Investigationsmentioning
confidence: 99%
“…In these versions the main attention is paid to implement parallel algorithm for covering by simplices. The efficiency of the developed parallel DISIMPL algorithm is investigated solving some standard test functions for global optimization and optimal design of grillage-type foundations problem [23].…”
Recently it was shown, that proposed simplicial partition based DISIMPL algorithm gives very competitive results to well known DIRECT algorithm for standard test functions and performs particularly well when the search space and the numbers of local and global optimizers may be reduced by avoiding symmetries. However, the simplicial partition has a cost (except for the case where the original feasible region is a simplex): in order to use simplicial partitions hyper-rectangular feasible region should be covered by simplices. After initial covering 2 n objective function evaluations is performed for DISIMPL-V and n! for DISIMPL-C. This limits DISIMPL method to rather small dimensional problems. To increase the applicability of the DISIMPL algorithm to higher dimensional problems a parallel DISIMPL for multicore computers was created and investigated in this paper. The efficiency of the developed parallel algorithm is investigated solving up to 55 dimensional optimal design of grillage-type foundations "blackbox" optimization problem.
“…The optimal scheme of grillage should possess the minimum possible number of piles. Theoretically, reactive forces in all piles should approach the limit magnitudes of reactions for them [23]. This goal can be achieved by choosing appropriate pile positions.…”
Section: Experimental Investigationsmentioning
confidence: 99%
“…In these versions the main attention is paid to implement parallel algorithm for covering by simplices. The efficiency of the developed parallel DISIMPL algorithm is investigated solving some standard test functions for global optimization and optimal design of grillage-type foundations problem [23].…”
Recently it was shown, that proposed simplicial partition based DISIMPL algorithm gives very competitive results to well known DIRECT algorithm for standard test functions and performs particularly well when the search space and the numbers of local and global optimizers may be reduced by avoiding symmetries. However, the simplicial partition has a cost (except for the case where the original feasible region is a simplex): in order to use simplicial partitions hyper-rectangular feasible region should be covered by simplices. After initial covering 2 n objective function evaluations is performed for DISIMPL-V and n! for DISIMPL-C. This limits DISIMPL method to rather small dimensional problems. To increase the applicability of the DISIMPL algorithm to higher dimensional problems a parallel DISIMPL for multicore computers was created and investigated in this paper. The efficiency of the developed parallel algorithm is investigated solving up to 55 dimensional optimal design of grillage-type foundations "blackbox" optimization problem.
“…The design parameters for both problems are the location of piles. An algorithm for local search was employed for the optimum location of piles beneath a se-parate beam of the grillage (Belevicius, Valentinavicius 2001, 2000 and under the whole grillage using an iterative algorithm on the basis of the above mentioned work (Belevicius et al 2002). Experience demonstrates that the objective function possesses many local minima points for practical grillage optimization problems.…”
The purpose of the paper is to present technology applied for the global optimization of grillage-type pile foundations (further grillages). The goal of optimization is to obtain the optimal layout of pile placement in the grillages. The problem can be categorized as a topology optimization problem. The objective function is comprised of maximum reactive force emerging in a pile. The reactive force is minimized during the procedure of optimization during which variables enclose the positions of piles beneath connecting beams. Reactive forces in all piles are computed utilizing an original algorithm implemented in the Fortran programming language. The algorithm is integrated into the MatLab environment where the optimization procedure is executed utilizing a genetic algorithm. The article also describes technology enabling the integration of MatLab and Fortran environments. The authors seek to evaluate the quality of a solution to the problem analyzing experimental results obtained applying the proposed technology.
“…The optimal scheme of grillage should possess the minimum possible number of piles. Theoretically, reactive forces in all piles should approach the limit magnitudes of reactions for those piles [5]. These goals can be achieved by choosing appropriate pile positions.…”
Section: Case Study: Optimization Problems Of Grillage-type Foundationsmentioning
Abstract. Branch and bound methods for global optimization are considered in this paper. Advantages and disadvantages of simplicial partitions for branch and bound are shown. A new general combinatorial approach for vertex triangulation of hyper-rectangular feasible regions is presented. Simplicial partitions may be used to vertex triangulate feasible regions of non rectangular shape defined by linear inequality constraints. Linear inequality constraints may be used to avoid symmetries in optimization problems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.