中文摘要：

在面向大规模业务的分布式服务流程应用环境中,服务流程的分布式执行引擎的自动配置管理是一个非常重要的问题。现阶段,简单的、静态的引擎节点配置并不能达到很好的效果。为此,提出了一种新的引擎自动配置管理算法,即基于模糊控制的服务流程引擎自动配置算法,同时在一个基于JINI的分布式服务流程引擎系统中,比较了该动态配置算法和常见的静态算法的性能。实验结果表明,该自动配置策略较静态策略而言能明显提高服务流程引擎系统整体响应时间的稳定性,使得服务流程引擎服务器能够根据外部请求负载情况进行自动的调节和配置,从而更好地适应不断变化的运行环境。

英文摘要：

Automatic configuration management of distributed executive engine in service-flow is essential to the applicable environment of large-scale business-oriented distributed service-flow. However, currentsimple and static configuration mechanism of the engine nodes doesn＇t work well. Therefore, a new automatic configuration management algorithm for distributed service-flow engine based on fuzzy control is introduced in this paper. And what＇s more, this dynamic configuration algorithm and the common static one were both implemented in a distributed service-flow engine system （SFES） based on the JINI platform to make a comparison on their performances. The experiment shows, this automatic mechanism has better performances than the static one, which remarkably improves the stability of overall response time of the SFES, and the service-flow engine server can automatically adjust and configure based on external requests of the load condition, so as to better adapt the changing operation environment.