Study On The Modelling And Evaluation Methods For High-speed Railway Signaling System Reliability And Availability

China's High-speed Rail has a word-leading,mature and complete technology system,which makes its operating mileage and construction scale rank first in the world.“The Belt and Road” strategy will further promote the rapid development of the world high-speed railway.The Medium-and Long-Term Railway Plan expands the Chinese network structure to“eight vertical and eight horizontal” corridors.The high-speed railway signaling system plays an important role to guarantee the efficient and safe operation of infrastructures and trains,where the reliability,availability and safety must be satisfied.As a safety-critical system,the high-speed railway signaling has two purposes: to avoid too high speed and to prevent violating a red signal.A failure occurring in a signaling system may have the potential to cause serious accidents such as train collisions and derailments.Currently,significant changes are taken places in the design,construction and operation phases,which makes the high-speed railway signaling system into a complex system with complex structure,high integration and large-scale equipment.As a result of the increasement of operating mileage and equipment quality,the number of facilities and the degree of association between the subsystems are increasing and the system reliability and availability evaluation are facing great challenges.Meanwhile,reliability and availability evaluation of high-speed railway signaling system is significant to the safety assessment and the maintenance improvement.Therefore,researches on the reliability and availability evaluations of signaling systems and equipment are of great important for the efficient and safe operation of high-speed railways.With the theory and method of mathematical statistics,Bayesian networks,"system of systems" and fuzzy set,the thesis develops the reliability and availability modeling and evaluation methods of high-speed railway signaling system.Based on the RAMS(reliability,availability,maintainability and safety)standards,the parameters related to high-speed railway signal systems are analyzed.Based on field data and expert experience,the characteristics of maintenance data and statistical property are utilized to obtain basic reliability data such as failure rates and repair rates.The main research contents are summarized as follows.1.Based on the field data and operation scenario,the reliability and availability of TCC(Train Control Center)and CTCS-3(Chinese Train Control System Level 3)onboard system are modeled by using the dynamic Bayesian network.With taking into account common cause failures,dynamic failure behaviors,imperfect coverage factors,and uncertainties on failure rates and repair rates in operational phase,the forward and backward inferences are conducted not only to evaluate and improve the reliability and availability,but also recognize the vulnerabilities of the onboard systems.Moreover,the models are compared for CTCS-3onboard systems with different redundant strategies,which can guide the choice of redundant structure in practice.The results obtain by the proposed model are validated with the field data from one railway bureau and maintenance decisions are provided.2.Based on superiority of universal generating function and dynamic Bayesian network for system multi-state modeling,the multi-state reliability and availability of high-speed railway signaling system,caused by transient,intermittent and permanent faults,are modeled.Moreover,with taking into account perfect repairs,imperfect repairs and condition-based maintenances,the multi-state reliability and availability of CTCS-3 onboard system is evaluated and compared.The influence of degradation probability on the performance of CTCS-3 onboard system is researched,which provides a theoretical foundation for multi-state reliability and availability assessment of high-speed railway signaling systems.3.Take advantage of “system of systems” theory,the models for overall reliability and availability of high-speed railway signaling systems are constructed to analyze the interaction mechanism and the cascade failures between constituent systems.Based on the Bayesian network,“system of systems” and multi-layer network,a hierarchical Bayesian network for the overall reliability and availability evaluation is proposed.The effects of failure rates,repair rates and imperfect coverage factors on the systemic reliability and availability are discussed to provide the overall improvement and standard establishment for high-speed railway signaling systems.Meanwhile,the model can realize the comprehensive trackside-onboard analysis for the reliability and availability evaluation.4.By combining the fault tree and fuzzy set theory,the overall reliability evaluation of high-speed railway signaling systems is proposed with taking into account the hardware factors,software factors,human factors and environment factors.The occurrence probabilities of known failure rate basic events are obtained from field data,while those of unknown failure rate basic events are estimated by expert judgment.The fuzzy set theory is then adopted to deal with the imprecise expert judgment.Considering a specific operational scenario,to accomplish one mission,the failure probability of CTCS-3 system is calculated.Importance analysis is carried out not only to identify the critical Basic events but also to provide suggestions on the improvement of preventive maintenance.Moreover,sensitivity analysis is conducted to illustrate the effect of hardware factors,software factors,human factors and environment factors on the reliability of CTCS-3 system.