Carrier Selection Optimization Based on Multi-commodity Reliability Criterion for a Stochastic Logistics Network under a Budget Constraint

In logistics management, selecting the carriers to deliver freight is a critical process for global enterprises. This paper determines the optimal carrier selection based on a multi-commodity reliability criterion for a logistics network subject to budget. Traditionally, a logistics network includes nodes and routes connecting the supplier and customer. Along each route, several carriers are available to deliver freight, which consists of multiple types of commodities. Since a carrier’s capacity for service may be reserved by other requests, every carrier will exhibit numerous possible capacities following a distinct probability distribution. Carrier selection must choose exactly one carrier for each route. Thus, any logistics network associated with a carrier selection is characterized as a multi-commodity stochastic flow network. We evaluate the probability that a network can satisfy the customer’s multi-commodity demand subject to a budget. This probability of multi-commodity reliability serves as a performance indicator for successful freight delivery. A genetic algorithm integrating minimal paths and Recursive Sum of Disjoint Products is developed to identify an optimal carrier selection strategy. A practical logistics network illustrates the computational efficiency of the proposed algorithm, comparing its performance with several algorithms.