基于广义随机Petri网的导弹系统测试性建模与指标评估方法研究

针对目前导弹测试性指标论证时缺乏有效方法支撑的问题，提出了一种基于广义随机Petri网（GSPN）的测试性建模方法。在建立导弹系统三维立体结构GSPN模型基础上，以各子系统的平面GSPN模型为研究对象，分析导弹系统故障检测及维修过程，同时根据国家军用标准GJB/Z1391—2006 故障模式影响及危害性分析指南对系统故障模式进行分类；基于GSPN模型可达标识图与嵌入马尔可夫链同构，以实存状态集和消失状态集表征库所变迁，采用同构法对测试性指标进行求解；以某型导弹子系统为例，对其测试性指标进行了评估，得到92%的故障检测率和90%的故障隔离率，验证了所提建模方法和相应算法的有效性，对测试性指标的权衡分析与评估具有指导意义。

Research on Evaluation Method for Testability Index and Modeling of Missile System Based on GSPN

A modeling method based on generalized stochastic petri net (GSPN) is proposed for the determination of missile testability indexes. Based on the GSPN model of three-dimensional structure of missile system, the planar GSPN model of each subsystem is taken as the research object, and the fault detection and maintenance process of missile system are analyzed. The missile system failure mode is classified according to the national military standard. The reachable marking graph based on GSPN model is isomorphic to the embedded Markov chain, the real state set and the disappeared state set are used to characterize the change of the library, and the isomorphic method is used to solve the testability indicators. Taking a certain missile subsystem as an example, the testability indicators were evaluated, and 92% of fault detection rate and 90% of isolation rate were obtained. The effectiveness of the proposed modeling method and corresponding algorithms are verified by using the testability indicators.