MC_IS {TesiproV}

MonteCarlo Simulation with importance sampling

Description

Method to calculate failure probability for structural engineering using a simulation method with importance sampling (a method to reduce the amount of needed samples)

Usage

MC_IS(
  lsf,
  lDistr,
  cov_user = 0.05,
  n_batch = 16,
  n_max = 1e+06,
  use_threads = 6,
  sys_type = "parallel",
  dataRecord = TRUE,
  beta_l = 100,
  densityType = "norm",
  dps = NULL,
  debug.level = 0
)

Arguments

lsf

objective function with limit state function in form of function(x) x[1]+x[2]...

lDistr

Distributions in input space

cov_user

The Coefficent of variation the simulation should reach

n_batch

Size per batch for parallel computing

n_max

maximum of iteration the MC should do - its like a stop criterion

use_threads

determine how many threads to split the work (1=singlecore, 2^n = multicore)

sys_type

Determine if parallel or serial system (in case MCIS calculates a system)

dataRecord

If True all single steps are recorded and available in the results file afteron

beta_l

In Systemcalculation: LSF´s with beta higher than beta_l wont be considered

densityType

determines what distributiontype should be taken for the h() density

dps

Vector of design points that sould be taken instead of the result of a FORM analysis

debug.level

If 0 no additional info if 2 high output during calculation

Value

The results will be provided within a list with the following objects. Acess them with "$"-accessor

pf probablity of failure

pf_FORM probablity of failure of the FORM Algorithm

var variation

cov_mc coefficent of the monteCarlo

n_mc number of iterations done

Author(s)

(C) 2021 - K. Nille-Hauf, T. Feiri, M. Ricker - Hochschule Biberach, Institut fuer Konstruktiven Ingenieurbau

References

DITLEVSEN O, MADSEN H. Structural reliability methods, vol. 178. New York: Wiley; 1996.

Spaethe, G.: Die Sicherheit tragender Baukonstruktionen, 2. Aufl. Wien: Springer, 1991. – ISBN 3-211-82348-4