项目名称: 基于主体的可信网构软件开发和运行的关键技术研究
项目编号: No.61303129
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 自动化技术、计算机技术
项目作者: 张大鹏
作者单位: 燕山大学
项目金额: 25万元
中文摘要: 网构软件(Internetware)是在当前开放、动态、难控的网络计算环境下产生的新的软件形态。如何在网构软件的开发和运行的过程中保障网构软件的可信性已经成为研究者们关注的热点问题。本课题拟对基于主体的可信网构软件的开发和运行中的关键技术进行研究。首先,在模型检测方面,针对已有的适用于基于主体的网构软件的建模语言与模型检测工具输入语言不匹配的问题,研究二者间自动/半自动的转换方法,给出相关算法并实现相关工具;其次,在信任动态预测方面,建立一种基于推荐偏差的网构软件信任模型,给出信任动态预测算法并进行仿真实验;最后,在故障诊断方面,建立基于主体的网构软件的故障诊断模型,研究使用动态描述逻辑DDL对诊断模型进行形式化描述的方法,给出相应的诊断算法,开发诊断工具并进行仿真实验。上述研究将从理论、模型和技术上为基于主体的可信网构软件的开发提供可信保障。
中文关键词: 模型检测;推荐偏差;BDI;最大频繁项目集;自动编码机
英文摘要: Internetware is a novel software paradigm under the currently open, dynamic, and uncertain network computing environment. Researchers have begun to focus on how to guarantee the trustworthy of internetware during its developing and running. We plan to carry out our research on key issues of developing and running of agent based trusted internetware. Firstly, in model checking, due to the mismatch of the current modeling language and the language that model checking tool uses, we research the automatic/semi-automatic transition method between them and realize the corresponding algorithms and tools. Secondly, in dynamic trust prediction, we build a internetware trust model based on recommend deviation. Then we give the algorithms of dynamic trust prediction and make simulation experiment. Finally, in fault diagnose, we build a fault diagnose model for agent based internetware, describe formally the model with dynamic description language (DDL). Then we give the corresponding algorighms, develop diagnose tool and make simulation experiment. The above research will provide guarantee for agent based trusted internetware from theories, modeling and technologies.
英文关键词: model checking;recommend deviation;BDI;maximal frequent itemsets;auto-encoder