基于Netica的导弹故障诊断贝叶斯网络模型研究

针对传统导弹装备故障诊断方法效率低、精确度差等问题,提出了采用贝叶斯网络进行导弹测试故障诊断的方法。本文首先对贝叶斯网络进行了概述,然后对导弹模型故障分支进行了贝叶斯网络故障诊断精确推理。进而采用故障树-贝叶斯网络转化法生成导弹典型故障贝叶斯网络模型,并使用贝叶斯工具Netica对该模型进行了案例分析。结果表明该方法可为导弹装备故障的快速定位提供依据,为贝叶斯网络技术在导弹故障诊断领域的应用奠定基础。

Research on Bayesian Network for Fault Diagnosis of Tactical Missile Using Netica

Aiming at the low efficiency and bad accuracy of traditional fault diagnosis of the tactical missiles,an alternative fault diagnosis method for missile test using Bayesian network is proposed.The concept of Bayesian network is briefly introduced at first,followed by precise reasoning of Bayesian network for a branch of the simplified model of the missile.Then a relevant Bayesian network model is established based on the common faults of the tactical missile by transforming fault tree to Bayesian network,and further reasoning is demonstrated with a simulating tool named Netica.The results of simulation prove that the method offers evidence for fast fault location of missiles,and provides a basis for the application of Bayesian network in the field of missile fault diagnosis.