Testability design is an effective way to realize the fault detection and isolation.Its important step is to determine testability figures of merits(TFOM).Firstly,some influence factors for TFOMs are analyzed,such as ...Testability design is an effective way to realize the fault detection and isolation.Its important step is to determine testability figures of merits(TFOM).Firstly,some influence factors for TFOMs are analyzed,such as the processes of system operation,maintenance and support,fault detection and isolation and so on.Secondly,a testability requirement analysis model is built based on generalized stochastic Petri net(GSPN).Then,the system's reachable states are analyzed based on the model,a Markov chain isomorphic with Petri net is constructed,a state transition matrix is created and the system's steady state probability is obtained.The relationship between the steady state availability and testability parameters can be revealed and reasoned.Finally,an example shows that the proposed method can determine TFOM,such as fault detection rate and fault isolation rate,effectively and reasonably.展开更多
In this paper, data streams are classified into four types conforming to a standardized infrastructure of communication networks for a substation automation system （SAS） based on IEC61850 system. The data exchanged ...In this paper, data streams are classified into four types conforming to a standardized infrastructure of communication networks for a substation automation system （SAS） based on IEC61850 system. The data exchanged on the net are demonstrated to be stochastic according to investigation on the Ethemet communication principles. Four generalized stochastic Petri nets （GSPN） based models for performance analysis of communication networks of IEC61850 system are developed based on the three-level structure of SAS, different time requirements of the four data streams and different networks topology for different voltage level. The GSPN-based model associated with immediate and exponential transitions is proven to be theoretically isomorphic with Markov chain; hence we apply the mathematic methods of performance evaluation contained in Markov chain to the GSPN models proposed. The computer simulation of the model including only sample value data streams shows that it can meet performance evaluation needs of communication networks of IEC61850 system. Further researches should be focused on the pe^ormance of the other three models to explain clear how those different data streams are interrelated to and interact on each other.展开更多
提出一种基于广义随机Petri网(Generalized Stochastic Petri Net,GSPN)的舰载服务器系统可靠性分析方法。深入分析舰载服务器系统故障模式的基础上,建立冗余服务器子系统的GSPN模型和共享数据盘子系统的GSPN模型,进而综合得到舰载服务...提出一种基于广义随机Petri网(Generalized Stochastic Petri Net,GSPN)的舰载服务器系统可靠性分析方法。深入分析舰载服务器系统故障模式的基础上,建立冗余服务器子系统的GSPN模型和共享数据盘子系统的GSPN模型,进而综合得到舰载服务器系统的全局GSPN模型,有效模拟了舰载服务器系统的动态行为。仿真实验验证了所提方法的有效性,为舰载服务器系统的分析与设计提供理论参考。展开更多
以广义随机Petri网(Generalized Stochastic Petri Net,GSPN)为数学工具,提出一种基于GSPN的计算机硬件系统可靠性建模与分析方法.在给出GSPN形式化定义和模型要素图形化表示和用法的基础上,建立了计算机硬件系统基本单元GSPN模型,该模...以广义随机Petri网(Generalized Stochastic Petri Net,GSPN)为数学工具,提出一种基于GSPN的计算机硬件系统可靠性建模与分析方法.在给出GSPN形式化定义和模型要素图形化表示和用法的基础上,建立了计算机硬件系统基本单元GSPN模型,该模型适用于描述结构简单的计算机硬件系统,而对于逻辑结构关系复杂的计算机硬件系统,由于其建模粒度过于精细势必造成组合爆炸问题.为此,对该计算机硬件系统基本单元GSPN模型进行精化设计,得到的精化模型可以用于描述结构复杂的计算机硬件系统.实例研究验证了所提方法的有效性,可以为计算机硬件系统可靠性设计提供理论指导.展开更多
对Petri网理论在性能评价领域的应用进行了研究,在介绍广义随机Petri网(GSPN,generalized stochastic petri net)及其分析技术的基础上,探讨了如何利用GSPN对系统进行建模及性能分析。针对典型的双处理机系统的吞吐量和资源利用率问题,...对Petri网理论在性能评价领域的应用进行了研究,在介绍广义随机Petri网(GSPN,generalized stochastic petri net)及其分析技术的基础上,探讨了如何利用GSPN对系统进行建模及性能分析。针对典型的双处理机系统的吞吐量和资源利用率问题,建立了双处理机系统的GSPN模型,利用随机Petri网与马尔可夫链的同构关系,得到了GSPN模型稳定的状态空间及各状态出现的概率,进而探讨了双处理机系统的吞吐量和利用率等性能指标。通过TimeNet工具对所建立的模型进行仿真实验,验证了模型的有效性,为吞吐量和利用率的提高提供相关依据。展开更多
使用广义随机Petri网(generalized stochastic Petri net,GSPN)对具有复杂时间约束信息的大规模Web服务组合系统进行建模时,模型难以直观理解,并且会造成状态空间爆炸.针对此问题提出一种扩展GSPN建模方法.将不同类型的时间约束信息分...使用广义随机Petri网(generalized stochastic Petri net,GSPN)对具有复杂时间约束信息的大规模Web服务组合系统进行建模时,模型难以直观理解,并且会造成状态空间爆炸.针对此问题提出一种扩展GSPN建模方法.将不同类型的时间约束信息分别施加于模型的位置、变迁和弧,可以简洁直观地表达系统复杂时间约束信息.进而提出一组应用于扩展模型的化简规则,对模型中一些常用结构进行化简,克服了GSPN缺乏通用化简方法以及难以在保留时间约束特征前提下进行化简的问题,减小了系统模型的状态空间.多种模型下的化简示例表明,所提方法有效降低了系统性能分析的复杂度,适用于对大多数Web服务组合系统进行快速性能分析.展开更多
文摘Testability design is an effective way to realize the fault detection and isolation.Its important step is to determine testability figures of merits(TFOM).Firstly,some influence factors for TFOMs are analyzed,such as the processes of system operation,maintenance and support,fault detection and isolation and so on.Secondly,a testability requirement analysis model is built based on generalized stochastic Petri net(GSPN).Then,the system's reachable states are analyzed based on the model,a Markov chain isomorphic with Petri net is constructed,a state transition matrix is created and the system's steady state probability is obtained.The relationship between the steady state availability and testability parameters can be revealed and reasoned.Finally,an example shows that the proposed method can determine TFOM,such as fault detection rate and fault isolation rate,effectively and reasonably.
文摘In this paper, data streams are classified into four types conforming to a standardized infrastructure of communication networks for a substation automation system （SAS） based on IEC61850 system. The data exchanged on the net are demonstrated to be stochastic according to investigation on the Ethemet communication principles. Four generalized stochastic Petri nets （GSPN） based models for performance analysis of communication networks of IEC61850 system are developed based on the three-level structure of SAS, different time requirements of the four data streams and different networks topology for different voltage level. The GSPN-based model associated with immediate and exponential transitions is proven to be theoretically isomorphic with Markov chain; hence we apply the mathematic methods of performance evaluation contained in Markov chain to the GSPN models proposed. The computer simulation of the model including only sample value data streams shows that it can meet performance evaluation needs of communication networks of IEC61850 system. Further researches should be focused on the pe^ormance of the other three models to explain clear how those different data streams are interrelated to and interact on each other.
文摘提出一种基于广义随机Petri网(Generalized Stochastic Petri Net,GSPN)的舰载服务器系统可靠性分析方法。深入分析舰载服务器系统故障模式的基础上,建立冗余服务器子系统的GSPN模型和共享数据盘子系统的GSPN模型,进而综合得到舰载服务器系统的全局GSPN模型,有效模拟了舰载服务器系统的动态行为。仿真实验验证了所提方法的有效性,为舰载服务器系统的分析与设计提供理论参考。
文摘以广义随机Petri网(Generalized Stochastic Petri Net,GSPN)为数学工具,提出一种基于GSPN的计算机硬件系统可靠性建模与分析方法.在给出GSPN形式化定义和模型要素图形化表示和用法的基础上,建立了计算机硬件系统基本单元GSPN模型,该模型适用于描述结构简单的计算机硬件系统,而对于逻辑结构关系复杂的计算机硬件系统,由于其建模粒度过于精细势必造成组合爆炸问题.为此,对该计算机硬件系统基本单元GSPN模型进行精化设计,得到的精化模型可以用于描述结构复杂的计算机硬件系统.实例研究验证了所提方法的有效性,可以为计算机硬件系统可靠性设计提供理论指导.
文摘对Petri网理论在性能评价领域的应用进行了研究,在介绍广义随机Petri网(GSPN,generalized stochastic petri net)及其分析技术的基础上,探讨了如何利用GSPN对系统进行建模及性能分析。针对典型的双处理机系统的吞吐量和资源利用率问题,建立了双处理机系统的GSPN模型,利用随机Petri网与马尔可夫链的同构关系,得到了GSPN模型稳定的状态空间及各状态出现的概率,进而探讨了双处理机系统的吞吐量和利用率等性能指标。通过TimeNet工具对所建立的模型进行仿真实验,验证了模型的有效性,为吞吐量和利用率的提高提供相关依据。
文摘使用广义随机Petri网(generalized stochastic Petri net,GSPN)对具有复杂时间约束信息的大规模Web服务组合系统进行建模时,模型难以直观理解,并且会造成状态空间爆炸.针对此问题提出一种扩展GSPN建模方法.将不同类型的时间约束信息分别施加于模型的位置、变迁和弧,可以简洁直观地表达系统复杂时间约束信息.进而提出一组应用于扩展模型的化简规则,对模型中一些常用结构进行化简,克服了GSPN缺乏通用化简方法以及难以在保留时间约束特征前提下进行化简的问题,减小了系统模型的状态空间.多种模型下的化简示例表明,所提方法有效降低了系统性能分析的复杂度,适用于对大多数Web服务组合系统进行快速性能分析.
