TY - GEN
T1 - On the Reliability of Dynamical Stochastic Binary Systems
AU - Lagos, Guido
AU - Romero, Pablo
N1 - Funding Information:
Acknowledgment. This work is partially supported by Project ANII FCE 1 2019 1 156693 Teoría y Construcción de Redes de Máxima Confiabilidad, MATHAMSUD 19-MATH-03 Rare events analysis in multi-component systems with dependent components and STIC-AMSUD ACCON Algorithms for the capacity crunch problem in optical networks. Guido Lagos acknowledges the financial support of FONDECYT grant 3180767.
Publisher Copyright:
© 2020, Springer Nature Switzerland AG.
PY - 2020
Y1 - 2020
N2 - In system reliability analysis, the goal is to understand the correct operation of a multi-component on-off system, i.e., each component can be either working or not, and each component fails randomly. The reliability of a system is the probability of correct operation. Since the reliability evaluation is a hard problem, the scientific literature offers both efficient reliability estimations and exact exponential-time evaluation methods. In this work, the concept of Dynamical Stochastic Binary Systems (DSBS) is introduced. Samaniego signature provides a method to find the reliability and Mean-Time-to-Failure of a DSBS. However, we formally prove that the computation of Samaniego signature belongs to the hierarchy of # P -Complete problems. The interplay between static and dynamic models is here studied. Two methodologies for the reliability evaluation are presented. A discussion of its applications to structural reliability and analysis of dependent failures in the novel setting of DSBS is also included.
AB - In system reliability analysis, the goal is to understand the correct operation of a multi-component on-off system, i.e., each component can be either working or not, and each component fails randomly. The reliability of a system is the probability of correct operation. Since the reliability evaluation is a hard problem, the scientific literature offers both efficient reliability estimations and exact exponential-time evaluation methods. In this work, the concept of Dynamical Stochastic Binary Systems (DSBS) is introduced. Samaniego signature provides a method to find the reliability and Mean-Time-to-Failure of a DSBS. However, we formally prove that the computation of Samaniego signature belongs to the hierarchy of # P -Complete problems. The interplay between static and dynamic models is here studied. Two methodologies for the reliability evaluation are presented. A discussion of its applications to structural reliability and analysis of dependent failures in the novel setting of DSBS is also included.
KW - Computational complexity
KW - Crude Monte Carlo
KW - Network optimization
KW - Reliability
KW - Samaniego signature
UR - http://www.scopus.com/inward/record.url?scp=85101199416&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-64583-0_46
DO - 10.1007/978-3-030-64583-0_46
M3 - Conference contribution
AN - SCOPUS:85101199416
SN - 9783030645823
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 516
EP - 527
BT - Machine Learning, Optimization, and Data Science - 6th International Conference, LOD 2020, Revised Selected Papers
A2 - Nicosia, Giuseppe
A2 - Ojha, Varun
A2 - La Malfa, Emanuele
A2 - Jansen, Giorgio
A2 - Sciacca, Vincenzo
A2 - Pardalos, Panos
A2 - Giuffrida, Giovanni
A2 - Umeton, Renato
PB - Springer Science and Business Media Deutschland GmbH
T2 - 6th International Conference on Machine Learning, Optimization, and Data Science, LOD 2020
Y2 - 19 July 2020 through 23 July 2020
ER -