CAJ | 학술논문

考虑政府采购部门以多阶段多属性逆向拍卖的形式将研发合约外包或采购某种需要研发的创新性产品。在每一阶段投标者的R＆D投入实现后，技术或产品的质量水平以一定的概率得以提高，采购方通过多属性拍卖来确定本阶段采购合同的获得者。投标者面临的问题是如何确定R＆D投入以最大化自己的累计净收益。构建一个随机动态博弈模型来研究多阶段多属性拍卖中投标者的策略，提供一个随机算法来发现随该机动态博弈的近似马尔科夫完美均衡，即投标者的最优R＆D投入策略。
고필정부채구부문이다계단다속성역향박매적형식장연발합약외포혹채구모충수요연발적창신성산품。재매일계단투표자적R＆D투입실현후，기술혹산품적질량수평이일정적개솔득이제고，채구방통과다속성박매래학정본계단채구합동적획득자。투표자면림적문제시여하학정R＆D투입이최대화자기적루계정수익。구건일개수궤동태박혁모형래연구다계단다속성박매중투표자적책략，제공일개수궤산법래발현수해궤동태박혁적근사마이과부완미균형，즉투표자적최우R＆D투입책략。
The paper studies a multi -stage multi -attribute reverse auction of R&D contracts or innovative products in government procurement.After R&D input is realized,the quality of potential bidders can be improved according to some probability.The winner of the contract in every stage is determined through a multi-attribute auction.The problem of the bidder is to determine the amount of R&D input to maximize his expected discounted value of future net cash flow.A sto-chastic dynamic game model is set up to analyze the strategy of the bidders and a stochastic algorithm is presented to find out Markov perfect equilibrium of the game.