An agent is a piece of code or a program that is capable of analyzing inputs from its environment to make decisions and then to take proper actions. This concept of rational agents was developed in the field of artificial intelligence being used to solve problems involving individuals when making decisions and taking actions as presented in studies of M. J. Mataric (1995) e P. Ciancarini, A. Omicini, and F. Zambonelli (2000).
The combination of several agents creates a multi-agent system in which complex interactions and behaviors are possible. Agents can communicate, collaborate, and compete with each other to achieve their goals. Some multi-agent systems feature a team of agents with a common goal. The UAV operation can surely be
considered a multi-agent system in which every aircraft is an agent with its own goals (destination, time frame of arrival, service standards, etc) and is independent of the goals of any other aircraft. With this approach, it is
possible to apply negotiation techniques that have been developed for such systems. These techniques will allow the aircraft to cooperate in solving airspace conflicts and, doing so, it will also be able to efficiently solve its own conflicts.
The possibility of cooperative negotiation is one of the powerful features of the multiagent systems. Agents are free to solve their conflicts by developing and implementing solutions together in a more efficient way than
when pursuing the solution individually. Agents will be free to communicate with one another and to negotiate among themselves, using a simple negotiation protocol known as the MCP - Monotonic Concession Protocol, to find safe and efficient solutions to conflicts, and to mitigate possible risks. Once the agents have agreed upon a resolution, the agents shall automatically conduct the negotiation process themseves,
excluding the participation of pilots or ground controllers.
0 comments:
Post a Comment