The Main Structure. Multi-agent Deal With The Environmental Mape, The Main Communication Language And The Main Constructor

Distributed Artificial Intelligence (DAI) is becoming a new hotspot of AI research. In DAI research, Multi-Agent System (MAS) is the most important. MAS can be seen as a system developed using bottom-up method, i.e., several autonomous agents are defined first, then we investigate how to solve a problem using one or more agents. MAS is the most suitable to describe the situation of human and human society. In order to investigate various problems in multi-agent systems, we proposed a Multi-Agent Processing Environment (MAPE) in 1994. MAPE is a developing, debugging, and running environment for multi-agent systems.The research work of this dissertation is focused on agent architecture, agent communication language, and agent construction paradigm. In this dissertation, I1. Summarize the origination of DAI and it's main research directions; Introduce MAPE briefly.2. Discuss several key topics of agent, including the definition of agent, current trends of the research of agent architecture and agent communication language.3. Propose the model of MAPE multi-agent system, the model of MAPE agent, and the compound architecture of MAPE agent. The compound architecture allows several intelligent components to coexists in an agent and run concurrently. They run independently, and interact with each other to achieve a common goal. Thus, an agent can react in time to emergencies, plan it's short- or middle-term behavior using certain strategy, model the environment and other agents to predict future events, and communicate with other agents to accomplish cooperation or coordination.4. Design an agent communication language SACL. SACL...