基于GSPN的导弹多层级测试性需求建模与指标评估

针对目前导弹测试性需求建模尚未成熟, 开展测试性需求分析难度较高的问题, 提出一种基于广义随机Petri网(generalized stochastic Petri net, GSPN)的测试性需求建模方法。在分析导弹结构的基础上对导弹测试性需求问题进行了建模, 给出导弹两级GSPN模型, 深入分析其导弹系统的故障检测及维修过程。根据GSPN模型同构于嵌入马尔可夫链(embedded Markov chain, EMC), 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。将多层级GSPN模型分解化简, 减少状态空间数量的求解。对导弹的GSPN模型进行定性和定量分析, 求解不同层级测试性指标最后通过仿真计算, 与系统要求可用度误差不超过0.7%, 验证了多层GSPN模型的可行性和分析结果的可信性。

Modeling and index evaluation of multi-level testability of missiles based on GSPN

Aiming at the problem that the test requirement modeling of missiles is not yet mature and it is difficult to conduct test requirement analysis,a test requirement modeling method based on generalized stochastic Petri net(GSPN)is proposed.Based on the analysis of the missile structure,the testability requirements of the missile are modeled,and a two-stage GSPN model of the missile is given to analyze the fault detection and maintenance process of the missile system.According to the GSPN model,it is isomorphic to the embedded Markov chain.The existing state set and disappearing state set are used to represent the change of the place,and the test index is solved by using the isomorphism method.The multi-level GSPN model is decomposed into multiple subnets from top to bottom,which greatly reduces the number of state spaces.The equivalent transition rule solves the compressed subnet.Finally,a combination of qualitative analysis and quantitative calculation is used to determine the test index requirements of missiles at all levels.Through simulation calculation,the availability error with the system requirements does not exceed 0.7%,which verifies the feasibility of the multi-layer GSPN model and the credibility of the analysis results.