A BDD & SEA based analysis of generalized phased-mission reliability

Guangyu Chen, Xizi Huang, Xiaowo Tang

DOI: https://doi.org/10.1109/RAMS.2006.1677416

2006-01-01

Abstract:Accounting for the reliability analysis of a generalized phased-mission system (GPMS) with two-level modular imperfect coverage (MIPC), the dynamic behavior & the statistical dependencies across phases are the most important factors for reliability modeling & analysis. There exist several approaches, such as a family of decision diagrams called ternary decision diagrams (TDD) and a multi-state binary decision diagram (MBDD) -based approach, proposed for the GPMS reliability evaluation. Based on simple & efficient algorithm (SEA) and binary decision diagram for phased-mission systems (PMS-BDD), we present a separable solution called GPMS-BDD approach, adding some new rules of phase algebra and a new BDD operation to ordinary rules of phase algebra and BDD operations of PMS-BDD. The computational complexity decreases due to taking advantage of the nature of the SEA & PMS-BDD, the separation of mission/phase imperfect coverage from the solution combinatorics & uses of phase algebra dealing with the dependencies across the phases. In this paper, the GPMS-BDD is presented, and compared with existing methods for analyzing the reliability of GPMS with MIPC & combinatorial phase requirements (CPR). An example generalized phased-mission system is analyzed to illustrate the advantage of our approach. GPMS analysis approach is an important part of reliability engineering. The upward two-level modular imperfect coverage model (UMIPCM) helps lower complexity of reliability analysis. The GPMS-BDD approach is efficient and integrated, helping a reliability engineer use existing software packages for fault-tolerant GPMS.