论文部分内容阅读
提出了一个具有高效问题求解能力的多Agent系统模型。基于生物免疫系统的自适应识别机制,将基因库进化与亲和度成熟、元动力学等要素相结合,设计具有自适应性和强大搜索能力的单Agent结构,多个Agent基于形态空间模型协同演化并涌现出智能求解能力。模型采用基于群体的多点随机搜索以及多Agent完全并行的执行方式,是一种新型协同演化模型。仿真实验结果证明了模型的有效性。
A multi-agent system model with efficient problem-solving ability is proposed. Based on the adaptive recognition mechanism of biological immune system, a single Agent structure with adaptive and powerful searching ability is designed by combining evolutionary and affinity maturation of genebank, elemental dynamics and so on. Many agents are based on morphological space model collaboration Evolution and emergence of intelligent solution. The model uses a group-based multi-point random search and multi-agent fully parallel execution mode, which is a new type of co-evolution model. Simulation results show the validity of the model.