Effect of ignoring input correlation on truck–shovel simulation

This paper presents an approach for handling correlated input variables in discrete event simulation (DES) modelling of truck–shovel systems using commercial DES software and uses a case study to investigate the effect of ignoring correlation between input variables. Multivariate random vectors, instead of independent probability distributions, are used for variables found to be correlated. The authors prove that correlations do exist in truck–shovel haulage systems. The model with multivariate random vectors performs better than the original model. The significance of modelling correlation in input variables depends on the strength of the correlation and the output’s sensitivity to the input variables.