Abstract: In the knowledge sharing system of industry-university-research collaborative innovation composed of universities and enterprises, a dynamic differential model based on double delay effect is constructed considering the influence of delayed effect on members’ knowledge sharing strategies. By using the maximum principle, the optimal knowledge sharing strategy, the optimal profit and the knowledge level of the system are obtained. The study finds that the existence of delay time has a negative effect on the knowledge sharing strategies of universities and enterprises, and is lower than the situation of immediate effects; Centralized decisions are not always better than decentralized ones. When the two delay times meet certain conditions, the centralized decision is the best choice; otherwise, the enterprises of both sides will pursue the maximization of their own interests and choose the decentralized decision. Under decentralized decision-making, only when the profit distribution ratio reaches a certain level will enterprises be willing to provide subsidies for colleges and universities. In addition, when the delay time of knowledge sharing in universities increases, enterprises will increase the subsidy proportion. The higher the proportion of subsidy, the higher the level of knowledge sharing in universities, but it has no influence on the strategy of knowledge sharing in enterprises.

Key words: industry-university-research, delay effect, collaborative innovation, knowledge sharing, dynamic differential model

摘要： 在高校和企业组成的产学研协同创新知识共享体系中,考虑成员的知识共享策略均存在延迟效应时的影响,构建了基于双延迟效应的动态微分模型。运用极大值原理,得到了协同创新主体在集中和分散两种决策下的最优知识共享策略、最优收益和体系的知识水平。研究发现:延迟时间的存在对高校和企业的知识共享策略具有负面效应,且均低于发生即时效应的情形;集中式决策并不总是优于分散式决策,两个延迟时间在满足一定的条件时,选择集中式决策为最优,否则双方企业将追求自身利益最大化而选择分散式决策;分散决策下,利润分配比例达到一定条件时,企业才愿意为高校提供补贴。且高校的知识共享延迟时间增加时,企业将提高该补贴比例。补贴比例越大,高校的知识共享水平越高,但对企业的知识共享策略没有影响。

关键词: 产学研, 延迟效应, 协同创新, 知识共享, 动态微分模型