中文摘要：

针对传统的BPMS处理异常时所采用的失败补偿法、ECA法、案例推理法等方案在智能化方面存在的不足，着重研究了业务流程管理系统中异常处理的问题，提出了一种ECA法与案例推理法相结合的、基于流程异常库的处理方式，来提高系统对于异常的预测能力以及在异常发生时能够尽早采取有效的解决措施，减小异常的负面影响，最大程度上实现系统处理异常的自动化及智能化，增加了BPMS的柔性。

英文摘要：

In allusion to the weakness of traditional methods BPMS used to deal with abnormity,including compensation of failure, ECA and case based reasoning,researches on the abnormity resolving in BPMS and brings forth a method based on process abnormity set,which combines ECA and CBR. It improves the forecasting capability of BPMS and makes sure to solve abnormity effectively as soon as possible, descend abnormity＇s negative affection, realize solving abnormity automatically and intelligently and then make BPMS more flexible.