TY - JOUR
T1 - A multiphase dynamic Bayesian networks methodology for the determination of safety integrity levels
AU - Cai, Baoping
AU - Liu, Yu
AU - Fan, Qian
PY - 2016/6
Y1 - 2016/6
N2 - A novel safety integrity levels (SILs) determination methodology based on multiphase dynamic Bayesian networks (MDBNs) for safety instrumented systems is proposed. Proof test interval phase and proof test phase are modeled separately using dynamic Bayesian networks, and integrated together to form the MDBNs. The unified structure models of MDBNs for k-out-of-n architectures are constructed, and the procedures of automatic creation of conditional probability tables are developed. The target failure measures, that is, probability of failure on demand, average probability of failure on demand, probability of failing safely, average probability of failing safely, and SIL of safety instrumented systems operating in a low demand mode, are evaluated using the proposed MDBNs. The effects of time interval of MDBNs, common cause weight, imperfect proof test and repair on model precision are researched. User-friendly SIL determination software is developed by using MATLAB GUI to assist engineers in determining the SIL value.
AB - A novel safety integrity levels (SILs) determination methodology based on multiphase dynamic Bayesian networks (MDBNs) for safety instrumented systems is proposed. Proof test interval phase and proof test phase are modeled separately using dynamic Bayesian networks, and integrated together to form the MDBNs. The unified structure models of MDBNs for k-out-of-n architectures are constructed, and the procedures of automatic creation of conditional probability tables are developed. The target failure measures, that is, probability of failure on demand, average probability of failure on demand, probability of failing safely, average probability of failing safely, and SIL of safety instrumented systems operating in a low demand mode, are evaluated using the proposed MDBNs. The effects of time interval of MDBNs, common cause weight, imperfect proof test and repair on model precision are researched. User-friendly SIL determination software is developed by using MATLAB GUI to assist engineers in determining the SIL value.
KW - KooM architecture
KW - KooMD architecture
KW - Multiphase dynamic bayesian networks
KW - Safety instrumented system
KW - Safety integrity level
UR - http://www.scopus.com/inward/record.url?scp=84958160129&partnerID=8YFLogxK
UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-84958160129&origin=recordpage
U2 - 10.1016/j.ress.2016.01.018
DO - 10.1016/j.ress.2016.01.018
M3 - RGC 21 - Publication in refereed journal
SN - 0951-8320
VL - 150
SP - 105
EP - 115
JO - Reliability Engineering and System Safety
JF - Reliability Engineering and System Safety
ER -