Application of genetic algorithm for logistics based on multi-agent system

Logistics is a crucial ingredient in the operation of an organization. Here how to efficiently deliver the supplies to the consumers in the field under the given condition is the key issue. The general logistics system is usually too complex to be modeled mathematically, or the models are overly computation intensive to be applied in real-time environment. This paper proposes a new method for logistics based on multi-agent system using modified A* algorithm to find the best route between the agents. In addition, genetic algorithm and K-medoids algorithm are used for selecting and clustering the agents to maximize the throughput of logistics. Computer simulation verifies the efficiency of the proposed scheme for a complicated application domain of military logistics.