Two stochastic optimization algorithms applied to nuclear reactor core design

Sacco, Wagner F.; Oliveira, C.R.E. de; Pereira, Cláudio M.N.A.

doi:10.1016/j.pnucene.2005.10.004

Cited by 54 publications

(31 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Regarding the turbine balancing problem, as the PCA was conceived as a continuous optimisation algorithm (Sacco and Pereira, 2006), first, we need to adapt it for combinatorial optimisation. In order to do so, we employ a representation technique named random keys (Bean, 1994).…”

Section: Implementation and Set-upmentioning

confidence: 99%

“…The Particle Collision Algorithm (PCA) (Sacco and de Oliveira, 2005;Sacco and Pereira, 2006;Rios-Coelho et al, 2010) is a Metropolis-based algorithm (Metropolis et al, 1953) that was introduced as an alternative to simulated annealing (Kirkpatrick et al, 1983). The main motivation behind the PCA was that in spite of being very powerful, the canonical simulated annealing is too sensitive to the choice of free parameters, such as, for example, the annealing schedule and initial temperature (Carter, 1997).…”

Section: Introductionmentioning

confidence: 99%

“…On the other hand, a particle that hits a low-fitness region is scattered to another region of the search space. The PCA, in its canonical form or variants, has been successfully applied to many realworld optimisation problems (Sacco and Pereira, 2006;Sacco et al, 2008;Knupp et al, 2009;Abuhamdah and Ayob, 2009;Rios-Coelho et al, 2010;Pacheco da Luz et al, 2011;Martinez González et al, 2014;Anochi and de Campos Velho, 2014).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A new Metropolis optimisation method, the cross-section particle collision algorithm: some preliminary results

Sacco

Rios-Coelho

2016

IJNEST

Self Cite

View full text Add to dashboard Cite

Abstract:In this work, we introduce a new Metropolis algorithm, which is an enhancement of the recent Particle Collision Algorithm (PCA), loosely inspired by the neutron interactions in a reactor. This novel method is called the CrossSection Particle Collision Algorithm (CSPCA), as it incorporates the concept of cross-section from Reactor Physics, in the sense that points in the search space and their respective fitness-function values are analogous to the neutron cross-sections which are used to express the likelihood of interaction between an incident neutron and a target nucleus. CSPCA is compared against the original PCA and two state-of-the-art metaheuristics, differential evolution and big bang-big crunch. These methods are applied to the turbine balancing problem, which is an NP-hard (i.e. non-deterministic polynomial-time hard) combinatorial optimisation problem that can be used to assess the potential of an algorithm to be applied to Fuel Management Optimisation (FMO). CSPCA performs better than its opponents, showing potential to be used not only in FMO, but also in other nuclear science and engineering optimisation problems.

show abstract

Section: Implementation and Set-upmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A new Metropolis optimisation method, the cross-section particle collision algorithm: some preliminary results

Sacco

Rios-Coelho

2016

IJNEST

Self Cite

View full text Add to dashboard Cite

show abstract

“…Some of these studies have focused on the representation of parameter uncertainty, either through specific probability functions, such as the Gaussian distribution (Bensal et al, 1998), or through the Poisson distribution (Orbán-Mihalyco et al, 2005) to solve these optimization problems. Other studies have used stochastic optimization to design process under uncertainty (Sacco et al 2006;Poplewski et al, 2011). Jamett et al (2012) applied stochastic programming for flotation design using a simple flotation model.…”

Section: Introductionmentioning

confidence: 99%

Solution Strategies to Stochastic Design of Mineral Flotation Plants

Jamett

Cisternas

Vielma

2014

Proceedings of the 8th International Conference on Foundations of Computer-Aided Process Design

View full text Add to dashboard Cite

The aim of this study is two-fold: first, to analyze the effect of stochastic uncertainty in the design of flotation circuits and second, to analyze different strategies for the solution of a two-stage stochastic problem applied to a copper flotation circuit. The paper begins by introducing a stochastic optimization problem whose aim is to find the best configuration of superstructure, equipment design and operational conditions, such as residence time and stream flows. Variability is considered in the copper price and ore grade. This variability is represented by scenarios with their respective probability of occurrence. The resulting optimization problem is a two-stage stochastic mixed integer nonlinear program (TS-MINLP), which can be extremely challenging to solve. For this reason, several solvers for this problem are compared and two stochastic programming methodologies are applied. The combination of these techniques allows the production of high quality solutions and an analysis of their sensitivity to epistemic uncertainty. The results show that the stochastic problem gives better designs because it allows operational parameters to adapt to the uncertainty of the parameters. The results also show that the flotation circuit structure can vary with the feed grade and copper price. The sensitivity analysis shows small to moderate variability with epistemically uncertain parameters.

show abstract

“…Primeiramente o problema inverso é formulado por meio do procedimento da máxima verossimilhança, resultando num problema de otimização (Moura Neto e Silva Schwaab e Pinto, 2007), onde a função objetivo é minimizada com os métodos estocásticos PCA (Particle Collision Algorithm) (Sacco et al, 2006) e DE (Differential Evolution) (Storn e Price, 1997;Price et al, 1 Introdução O processo de difusão de partículas é, em geral, modelado matematicamente pela clássica lei de Fick. No entanto, nesta formulação, alguns fenômenos que podem ocorrer juntamente com a difusão são completamente ignorados.…”

unclassified