A multi‐agent‐based negotiation support system for distributed transmission cost allocation

We have developed a multi‐agent system (MAS), based on the network flow model and KQML, called MASCAN, to support negotiations in the cost allocation of network transmission. This is very important to industries that have different entities connected with lines or pipes, such as the Internet and telecommunications. Such an approach is especially useful to the utility industries, such as electricity and gas, and the transportation industry. In the system, each agent represents a node in a network, for example supplier or consumer. Agents do not receive any centralized controls or information from centralized sources to guarantee autonomy–a key requirement for the agent. In this all decisions are made locally based on the rules or knowledge that each agent has or captured to communicate or coordinate with other agents for the cheapest path under fair‐play requirements. We also assume that each agent is rational, that is, one of the goals or objectives of agent decisions or movements is to minimize costs or increase profits. The solution to cost allocation is to search for the equilibrium point of a non‐cooperative game subject to the given constraints, for example network capacity. We applied MASCAN to model and support the negotiation of cost allocation in power transmission, and the results and how this approach supported the process of negotiation are perceived to be closer to the real‐world negotiation and the outcomes were accepted more easily by the participants. Copyright © 2001 John Wiley & Sons, Ltd.