Modeling Method Research Of Mission Reliability Of Equipment System Of Systems Based On DoDAF

The mission reliability of Equipment System of Systems(ESo S) means the capability to finish the mission under given mission profiles. Mission reliability of ESo S is important basis of generating and keeping combat capability, which influences the operation mode, scale and capability of sustainable combat. Besides, the ESo S has the characteristics of large scale, complex task, and dynamic structure. So, the reliability of ESo S is hard to describe and model. The research of mission reliability modeling methods can support the top-level design of ESo S, provides technology for ESo S design and optimization. The main works are as follows:(1) Building Do DAF-ROD models of ESo STo describe the ESo S reliability element, it is necessary to build the ESo S structure and mission description model. First, we analyze Do DAF(Department of Defense Architecture Framework) function, structure and problems of Do DAF. Then, aiming at the requirement of mission reliability, we choose 8 products from 52 products of Do DAF 2.0. Especially, we extend the 4 products to present the reliability logic relation, that is SV-4, OV-5a, OV-6c, SV-5a. At last, based on the development flow of Do DAF, we develop the Do DAF-ROD models of ESo S. We analyze the purpose and function of Do DAf-ROD 8 products, give the UML model and XML file of products. Do DAF-ROD is normalization description of mission reliability, which is the information input of computer aid building reliability model(2) Modeling rules and algorithm from Do DAF-ROD to BNAiming at compute aid generating and fast solving, it is necessary to research the compute aid generating rule and algorithm from Do DAF-ROD to ESo S reliability model. This paper gives the transformation rule and algorithm from 5 Do DAF-ROD products to Bayesian Networks. Operation activity decomposition tree is transferred to leaf nodes and middle nodes. The System interface description is transferred to root nodes. System function description model is transferred to condition probability table. System measurement is transferred to marginal probability. The Event trace description model is transferred to phase of mission and root nodes. The System function description is transferred to conditional probability table and non-root nodes. And finally we connect phase Bayesian Network to mission Bayesian Network. The Bayesian Network model can be used in the demonstration phase of ESo S.(3) Modeling rules and algorithm from Do DAF-ROD to MAMAiming at the dynamic characteristics of ESo S, we build Multi-Agent Model(MAM). This paper mainly focuses on the transforming rule and algorithm of 5 Do DAF-ROD products. The high level operation conceptual graph is transformed to scenario of Agent simulation, operation activity model is transformed to operation function of Agent. Operation State transition description is transformed to operation state, System State transition to reliability state and System Measurement matrix to reliability parameters. According to these steps, we obtain Multi-Agent simulation model. The MAM can be used in detailed design of ESo S.(4) Model validation and comparative analysisTo validate and compare different reliability model, this paper builds large and small scale scenarios. The paper gives Markov, BDD, Bayesian Network, Multi-agent 4 mission reliability model. The consistence of different model is validated by modeling missile defense scenario. The characteristics of different model is validated by modeling centralize missile defense system of systems and distribution missile defense system of systems.