| 摘要 |
PROBLEM TO BE SOLVED: To provide a multiagent system for balancing evaluation between agents and coordinating the agents with each other. SOLUTION: The agent 1 randomly selects a variable value x<SB>1</SB>as candidate data, acquires environmental data H(x<SB>1</SB>) from environment and uses an evaluation part 102 to calculate an evaluation value H<SB>1</SB>(x<SB>1</SB>). It requests the agent 2 to evaluate the variable value x<SB>1</SB>via a communication part 105 when a decision part 103 determines that the evaluation value increases. The agent 2 evaluates the variable value x<SB>1</SB>and notifies it of agreement if the evaluation value increases and claims the evaluation value for a decrement as a compensation if the evaluation value decreases. The agent 1 agrees with the agent 2 claiming the compensation when the evaluation value increases even if it accepts the compensation. With such a negotiation, the agents 1, 2 operate to make the evaluation value increase, whereby the candidate data is updated and output. COPYRIGHT: (C)2003,JPO
|
