Multiagent Systems in Logistics Environment
Abstract
A procurement problem for a manufacturing company is to decide the suppliers and manufacturing locations from a list of prospective candidates for the manufacture of products that will minimize the cost added to the price of the products. This is a typical logistics problem. Under some assumptions, this problem can be solved through negotiation in a forum of candidate transportation companies. From this perspective, we can model each transportation company as comprising of a master agent and several slave agents. A master agent negotiates for tasks with the master agents of other companies and assigns tasks to its own vehicles. A slave agent represents a vehicle which generates robust vehicle routes, and cost estimation for an assigned task by collaborating with other vehicles. We develop a theoretical information infrastructure for solving this problem using multi agents, game theory and stochastic programming. We apply this formalism to a logistics problem posed by USALIA and build an implementation platform.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 23, 2000
- Accession Number
- ADA379150
Entities
People
- R. T. Kumara
Organizations
- Pennsylvania State University