基于免疫机理的协同工作系统异常处理

Exception handling based on the immune mechanism in collaborative works

提出一种基于免疫机理的协同工作异常处理方法.它利用先验知识和事例学习获取的知识,归纳成知识库并将它们转化为免疫抗体库,将错误输入信息视为抗原,建立基于免疫机理的Agent模型,其中利用其对异常进行学习、记忆、抑制、恢复功能,并利用模糊量化的方法进行异常处理决策.该模型很好的利用了免疫系统的分布式、记忆特性、学习能力,能有效地对异常信息进行快速发现与处理.最后实现了一个原型系统,测试结果表明该方法是有效的.

A method based on the immune mechanism for handling exceptions in collaborative works is presented. Viewing fault input information as antigen library, it utilized knowledge learning from case study, induced them to knowledge and transformed to antibody library, and built an agent model based on immune mechanism to study, memorize, control exceptions and restore systems, in which decisions are made based on a fuzzy method. This model makes good use of the distribution, memorization, learning ability to discover and process the exceptional information quickly and effectively. Finally, a prototype system is implemented and the availability is shown by the result.