CAJ | 학술논문

通过构建基于知识观的原始性创新多Agent模型,剖析原始性创新支撑知识、发起与参与主体以及主体合作网络之间的联系,进而完成主体行为描述及其适应性规则设计,并且借助Swarm平台实施仿真.结果显示学科研究方向总数极小的系统,其原始性创新绩效最优.对于学科研究方向总数中等乃至稍多的系统,采用优先适应其他主体原始性创新的策略或者强强联合的策略都可以显著提升其原始性创新绩效,二者相比不存在绝对的最优.
통과구건기우지식관적원시성창신다Agent모형,부석원시성창신지탱지식、발기여삼여주체이급주체합작망락지간적련계,진이완성주체행위묘술급기괄응성규칙설계,병차차조Swarm평태실시방진.결과현시학과연구방향총수겁소적계통,기원시성창신적효최우.대우학과연구방향총수중등내지초다적계통,채용우선괄응기타주체원시성창신적책략혹자강강연합적책략도가이현저제승기원시성창신적효,이자상비불존재절대적최우.
It proposes a multi-agent model on the original innovation, and analyses the relationships among the supporting knowledge system, its participators, and participators collaboration network, design agents＇ behaviors as well as their adap- tion rules, and simulate the model on Swarm platform. It shows that the system with very few research directions tends to have high innovation performance. As for systems with middle or higher amount of research directions, adopting the strategy to cooperate with the original innovations proposed by others with priority and the strategy to induce cooperation between ex- perts may both improve the system innovation performance. These strategies both have their individual advantages, and can＇ t conclude the best one.