Design of Experiment Using Simulation of a Discrete Dynamical System

Abstract The topic of the presented paper is a promising approach to achieve optimal Design of Experiment (DoE), i.e. spreading of points within a design domain, using a simulation of a discrete dynamical system of interacting particles within an n-dimensional design space. The system of mutually repelling particles represents a physical analogy of the Audze-Eglājs (AE) optimization criterion and its periodical modification (PAE), respectively. The paper compares the performance of two approaches to implementation: a single-thread process using the JAVA language environment and a massively parallel solution employing the nVidia CUDA platform.

[1]  Miroslav Vořechovský,et al.  Improved formulation of Audze-Eglājs criterion for space-filling designs , 2015 .

[2]  Giacomo Gigante,et al.  On the Koksma-Hlawka inequality , 2011, J. Complex..

[3]  Miroslav Vorechovský,et al.  Modification of the Audze-Eglājs criterion to achieve a uniform distribution of sampling points , 2016, Adv. Eng. Softw..

[4]  Chang-Xing Ma,et al.  Wrap-Around L2-Discrepancy of Random Sampling, Latin Hypercube and Uniform Designs , 2001, J. Complex..

[5]  Kevin Skadron,et al.  A performance study of general-purpose applications on graphics processors using CUDA , 2008, J. Parallel Distributed Comput..

[6]  Miroslav Vorechovský,et al.  Hierarchical Refinement of Latin Hypercube Samples , 2015, Comput. Aided Civ. Infrastructure Eng..

[7]  A. Saltelli,et al.  Reliability Engineering and System Safety , 2008 .

[8]  Arnold J. Stromberg,et al.  Number-theoretic Methods in Statistics , 1996 .

[9]  Michael Chu,et al.  Scientific and Engineering Computing Using ATI Stream Technology , 2009, Computing in Science & Engineering.

[10]  A. Rollett,et al.  The Monte Carlo Method , 2004 .

[11]  S. Bates,et al.  Formulation of the Audze--Eglais uniform Latin hypercube design of experiments , 2003 .

[12]  D. Novák,et al.  CORRELATION CONTROL IN SMALL-SAMPLE MONTE CARLO TYPE SIMULATIONS I: A SIMULATED ANNEALING APPROACH , 2009 .

[13]  P. Frantík,et al.  Parallelization of lattice modelling for estimation of fracture process zone extent in cementitious composites , 2013, Adv. Eng. Softw..

[14]  Peter Winker,et al.  Centered L2-discrepancy of random sampling and Latin hypercube design, and construction of uniform designs , 2002, Math. Comput..

[15]  M. E. Johnson,et al.  Minimax and maximin distance designs , 1990 .

[16]  Ronald L. Iman,et al.  Risk methodology for geologic disposal of radioactive waste: small sample sensitivity analysis techniques for computer models, with an application to risk assessment , 1980 .

[17]  R. L. Iman,et al.  Latin hypercube sampling (program user's guide). [LHC, in FORTRAN] , 1980 .

[18]  Miroslav Vořechovský,et al.  Performance comparison of methods for design of experiments for analysis of tasks involving random variables , 2015 .