Abstract
The vehicle routing problem (VRP) determines a delivery route that minimizes the delivery cost. In this study, we consider the stochastic VRP with uncertainty and consider the variation in customer demand, which may cause a shortage of products during delivery. In this case, delivery vehicles have to return to the depot and replenish the products. We consider a model that minimizes the sum of the additional cost caused by the shortage and the normal delivery cost. In previous studies, the decomposition method using the L-shaped method was used. In this study, we improve the decomposition method to make it more efficient. In addition, we have improved the direct method of calculating the additional cost without the decomposition method by considering subtour elimination constraints. We have shown that the direct calculation method is superior in terms of time-saving.