Abstract: To improve service to replace the success rate of failure of cloud services collaboration process, it puts forward a collaborative situational evolution based adaptive candidate negotiation model based on cloud platform, uses Bayesian learning forecast consultation data retention value, adaptively submits negotiation issue, in order to maintain negotiate with candidate. Through dynamic monitoring process of service collaboration, it dynamically optimizes service candidates. The experimental results show that maintaining service consultation model can adapt to the collaborative process of change, make the service collaboration with good adaptability and robustness.

Key words: cloud service, collaboration, service failure, negotiation

摘要： 为提高云服务协作过程服务失效的替换成功率，提出一种云平台下协作态势驱动的自适应候选者协商维持模型，利用贝叶斯学习预测协商数据保留值，自适应提交协商议题，以维持与候选者的协商；通过对服务协作过程动态监测，动态优化服务候选集。实验结果表明，服务协商维持模型能适应协作过程变化，使服务协作具有良好的自适应性和健壮性。

关键词: 云服务, 协作, 服务失效, 协商