A cross-entropy heuristic statistical modeling for determining total stochastic material handling time

This paper concerns with proposing a heuristic statistical technique to compute total stochastic material handling time in an automated guided vehicle (AGV) equipped jobshop manufacturing system. With respect to stochastic times of AGVs material handling process, the material handling activities probability distributions are considered. Using the probability distributions, we model the AGV material handling problem using a heuristic statistical method when the activities’ probability distribution functions are the same. Also, in the case that the activities’ probability distribution functions are different, a cross-entropy approach is proposed and developed to model the problem. The effectiveness of the proposed model is illustrated in a numerical example and verified by a simulation study. The numerical experiments are worked out in two cases, namely having same probability distributions and different probability distributions for activities. Both cases are verified by simulation study using different simulation softwares showing the efficiency of the employed approaches. In case 1, the deviation of ARENA from the proposed statistical model is estimated to be 0.5 % while for other softwares detailed in the experiments the value of deviations are more than 30 %. On the other hand, in case 2 again ARENA performed better than others having the deviation of 0.3 % from the proposed cross-entropy.