Hjälp
Sök i LIBRIS databas

Sökning: onr:s2vt4px4qnhftjmp > Efficient use of Mo...

# Efficient use of Monte Carlo [Elektronisk resurs] The Fast Correlation Coefficient

Sjöstrand, Henrik, 1978- (författare)
4th edition of the International Workshop on Nuclear Data Covariances, 2017
Asquith, Nicola (författare)
Helgesson, Petter, 1986- (författare)
Rochman, Dimitri (författare)
van der Marck, Steven (författare)
Laboratory for Reactor Physics Systems Behaviour, Paul Scherrer Institut, Villigen, Switzerland (medarbetare)
Uppsala universitet Teknisk-naturvetenskapliga vetenskapsområdet (utgivare)
Engelska.
Ingår i: ; 4
Läs hela texten
Läs hela texten
Läs hela texten
Läs hela texten
• E-artikel/E-kapitel
Sammanfattning Ämnesord
Stäng
• Monte Carlo methods are increasingly used for Nuclear Data evaluation and propagation. In particular, the Total Monte Carlo (TMC) Method [1] has proved to be an efficient tool. A disadvantage of MC is that statistical uncertainties are also introduced. For evaluating the propagated nuclear data uncertainty, this was addressed with the so-called Fast-TMC method \cite{Rochman14}, which has become the standard route for TMC uncertainty propagation.Today, the standard way to visualize and interpret Nuclear Data (ND) co-variances is by the use of the Person correlation coefficient.$\rho = \frac{{{\mathop{\rm cov}} ({x},{y})}}{{{\sigma _{{x}}} \cdot {\sigma _{{y}}}}},$where x or y can be any parameter dependent on ND. As addressed in \cite{Rochman14}, $\sigma$ has both a ND component, $\sigma_{ND}$, and a statistical component, $\sigma_{stat}$. The contribution from $\sigma_{stat}$ decreases the value of $\rho$, and hence it is easy to underestimate the impact of the correlation. One way to address this is to minimize $\sigma_{stat}$ by using longer run-times. Alternatively, as proposed here, a so-called fast correlation coefficient is used,${\rho _{fast}} = \frac{{{\mathop{\rm cov}} (x,y) - {\mathop{\rm cov}} ({x_{stat}},{y_{stat}})}}{{\sqrt {\sigma _x^2 - \sigma _{x,stat}^2} \cdot \sqrt {\sigma _y^2 - \sigma _{y,stat}^2} }}$In many cases, ${\mathop{\rm cov}} ({x_{stat}},{y_{stat}})$ can be assumed to be zero.The paper explores two examples, correlations from the NRG High Flux Reactor spectrum \cite{Asquith16} and the correlations between different integral criticality experiments. The impact of the use of $\rho_{fast}$ is quantified, and the implication of the results are discussed.

### Ämnesord

Natural Sciences  (hsv)
Physical Sciences  (hsv)
Subatomic Physics  (hsv)
Naturvetenskap  (hsv)
Fysik  (hsv)
Subatomär fysik  (hsv)
Inställningar Hjälp

Beståndsinformation saknas

### Sök utanför LIBRIS

Hjälp
Om LIBRIS
Sekretess
Blogg
Hjälp
Fel i posten?
Kontakt
Teknik och format
Sök utifrån
Sökrutor
Plug-ins
Bookmarklet
Anpassa
Textstorlek
Kontrast
Vyer
LIBRIS söktjänster
SwePub
Sondera
Uppsök

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.