摘要
针对软件系统可靠性建模难问题,提出了一种新的适用于软件系统可靠性估计的模糊神经Petri网(简称为FN-PN)。文章首先给出了模糊神经Petri网的定义及其引发规则,然后给出了一种学习算法。该FNPN结合了模糊Petri网和神经网络各自的优点,既可以表示和处理模糊产生式规则的知识库系统又具有学习能力,可通过对样本数据学习调整模型中的参数以获得系统内部的等效结构,从而计算出非样本数据的系统的可靠度。最后用实例分析验证了该方法的可行性。
Aiming at the practical conditions of software systems,a fuzzy neural Petri-net method(FNPN) for estimating the reliability of software systems is proposed.The definition and fire-rules of this FNPN are presented.Then a learning algorithm is put forward.It can express and process the knowledge-based system of fuzzy productive rules,and possesses the capability of learning.Finally the model of reliability evaluation of information systems is established based on it.According to qualitative analysis and quantitative indices calculation,the results show that the model is effective and reasonable.
出处
《计算机与数字工程》
2010年第12期1-3,13,共4页
Computer & Digital Engineering
基金
国家自然科学基金项目(编号:30572374)资助
