摘要
在模型驱动的软件自适应控制过程中,监测、分析、决策和执行等活动均基于共享的知识模型。为便于知识重用和运行时维护,常采用抽象级别较高的需求模型来表示知识。为建模软件的适应性需求,针对传统的Tropos及其扩展方法不能用于软件对异常事件适应性需求建模问题,对Tropos进行上下文和异常条件扩展,记为Tropos+。在此基础上,提出一种由Tropos+需求模型驱动的软件自适应方法,该方法能够用于软件运行环境和异常事件监测以及软件对环境变化和异常事件的自适应处理。最后通过一个案例说明了软件自适应过程。
In the self-adaptive control process of model-driven software adaption, the actions of moni- toring, analyzing, planning and executing are based on the shared knowledge model. For the convenience of knowledge models＇ maintaining and reusing, highly abstract requirement models are usually used to represent the knowledge. In order to model the adaptive requirements of software and to solve the problem that traditional Tropos and its extended methods cannot model the software adaptive requirements to exceptional events,we propose the Tropos＋ which can monitor and deal with context changes and exception events. Base on Tropos＋ , we present a requirement model-driven software self-adaption method. Finally, the process of software self-adaption based on the proposed method is illustrated by an example.
出处
《计算机工程与科学》
CSCD
北大核心
2015年第10期1877-1883,共7页
Computer Engineering & Science
基金
软件工程国家重点实验室基金资助项目(SKLSE 2012-09-38)
关键词
自适应软件
需求模型
自适应控制过程
self-adaptive software
requirement model
adaptive control process
