Research On Automated Safety Analysis Method For Fully Automatic Operation System

With its advanced technology advantages,the fully automatic operation system has become the development trend of the global rail transit system.As a safety critical system,safety analysis is an indispensable part of it.The safety analysis of the rail transit system needs to be iteratively carried out With the increase of system complexity and integration,the workload of safety analysis is also significantly increased,which makes the safety analysis often lag behind the design and cannot provide suggestions for system design in time.In addition,traditional safety analysis relies on manual analysis by analysts,which is not only less efficient,but also increases the probability of making mistakes.Completing safety analysis in an automated way can meet the needs of complex systems for safety analysis,greatly improve the efficiency of safety analysis,reduce manual workload and human error,ensure the correctness and comprehensiveness of safety analysis,and keep synchronization with the system development process,so as to achieve a "safe-drive" design.Based on the System-Theoretic Process Analysis(STPA)method,this thesis proposes an automated safety analysis method for fully automatic operation system,which combines the technical characteristics of automatic operation system.The operational scenarios of Beijing Yanfang Line are analyzed by an example.The innovations of this thesis are as follows:(1)Aiming at the problem that there are too many layers in the hierarchical control structure diagram of STPA method,which makes it impossible to effectively trace the cause,and the timing problem of multiple control processes does not be considered,a basic control structure model is defined,which can model multiple control processes in time sequence and establish traceability relationship through variables.Then the timing-related accident causal scenarios can be identified,and the causal type of the accident is expanded.(2)Aiming at the problem that the basic control structure model only contains the control structural information and lacks the causal information,the basic control structure model is extended from the four aspects:control behavior,input variables,external interference and synchronization timing.More system causal information is included in the control model as the basis for generating unsafe control behavior and causal scenarios.In addition,the attributes of the control behavior in the fully automatic operation system are divided from three aspects:control mechanism,application time and the relationship with preconditions.The association matrix between the attributes of the control behavior and the unsafe control behavior is analyzed and established,so that the automatic identification of the unsafe control behavior can be completed.(3)As an important analytical output of the STPA method,the accident causal scenario can describe the process logic of the accident clearly and concretely from the perspective of the control mechanism.Only by improving the quality of the causal scenario can the effective control measures be taken.However,the STPA method lacks a unified standard description for the cause scenario,then a four-stage causal scenario description method is defined in this thesis,which includes the operational background,control defect,unsafe control behavior,and system-level accident.In addition,this thesis has developed the first timing,non-first timing,synchronization timing and external interference causal scenario search rules to ensure the automatic identification of the causal scenarios,besides,the dynamic of the causal scenarios are realized.Based on the automated safety analysis method proposed in the thesis,the Auto-STPA platform was designed and developed,which was applied to the operational scenarios of Beijing Yanfang Line,and successfully completed the automatic identification and dynamic display of related causal scenarios.Besides,the corresponding safety requirements are added.Finally,the correctness,comprehensiveness and efficiency of the method are fully verified.