Mean-variance approach to the management of cyclic deliveries

The scheduling problem of cyclic deliveries presented in this paper is characterized by a set of customers requiring, from a centralized supplier, stochastically varying quantities of a given commodity with prefixed time periodicity. The model is of the multiobjective type; the customers are scheduled in order to minimize both the mean value and the variance of the capacity of the distribution system. A heuristic algorithm to solve this problem in polynomial time is described and an experimental analysis of its performance is presented.