考虑需求与研发水平不确定性，以及研发竞争，构建需求服从带随机跳跃的几何布朗运动。在此基础上，运用实物期权和Stackelberg博弈建立领先者与追随者研发投资决策模型。通过对模型的研究表明：研发成功率影响博弈均衡。当领先者的平均研发成功率高于两者期权相切对应的成功率，或追随者的平均研发成功率低于两者期权相切对应的成功率，抢先均衡；否则，序贯均衡。领先者与追随者随着需求波动的增加，竞争者研发水平的提高或竞争者研发成功需求下降幅度的增加而推迟研发投资，随着自身研发水平的提高而提前研发投资。最后，采用一个算例验证了模型的有效性。

Considering the uncertainty of demand and R&D level, and R & D competition, we construct a geometric Brownian motion with random jump. On this basis, we use real options and Stackelberg game to establish a leader and follower R & D investment decision model. The research on the model shows that the R&D success rate affects the game equilibrium. When the average R&D success rate of the leader is higher than the corresponding success rate of the tangent of the two option values, or the average R&D success rate of the followers is lower than the corresponding success rate of the tangent of the two option values, there is a preemptive equilibrium, otherwise, there is a sequential equilibrium. Leader and follower will postpone R&D investment as the demand fluctuation increases, the competitor’s R&D level increases or the degree of declining in demand increases for R & D success of competitor, advance R&D investment as self R&D level increases. Finally, the validity of the model is verified empirically.

F2

适应复杂需求的SMPEs运营作业系统管理与优化研究（编号：71271060），订单式生产人工作业系统（MTO/MOS）组织与优化研究（编号：70971026）；广州市哲学社会科学“十三五”规划2018年度课题：基于科技金融大数据下科技企业价值评估与风险管理研究(编号：2018GZYB91)