多假设成立的多agent合作决策

The Algorithm of Multi-Agent Cooperative Decision-Making under Multiple Hypotheses Correct Synchronously

对复杂环境对象进行决策时,多agent合作可以融合各agent的知识经验,提高决策结果的可靠性.针对环境对象的辨识空间中多假设同时成立的决策问题,本文提出了一种基于证据理论的多agent合作决策算法,详细描述与分析了多agent合作决策的原理.为提高系统决策的可靠性,降低了合成计算的复杂度,在多agent合作决策系统中引入正确的训练案例进行学习.本文提出的算法应用于蔬菜的病害判别,实验结果验证了本文提出的多agent合作决策算法的有效性.

While Multiple agents cooperative decision-making for complex environment object, each agent's priori-knowledge is combined, so that the reliability of results is enhanced. This paper presented a multi-agent cooperative decision-making algorithm based on Dempster-Shafer theory for the objects whose frames of discernment exist multiple correct hypotheses synchronously. The theory of multi-agent cooperative decision-making is analyzed and described in detail. In order to enhance the decision-making reliability and decrease combinative computing complexity, the correct training patterns are introduced for the multi-agent system. This algorithm is applied to vegetable diseases recognition. The results demonstrate the multi-agent cooperative decision-making algorithm effective.