Research On Safety Resilience Theory And Method For Next Generation Air Transportation System

With the rapid growth of air traffic,the next generation air transportation system(Next Gen)is a high distribution,software-intensive,safety as the critical socio-technology system,which complexity and coupling are increasing.The application of cloud computing,big data,artificial intelligence,and other next-generation information technologies in the civil aviation industry is deepening.The air transportation system needs to use these new technologies to solve safe and efficient guarantee service in the complex operating environment.These new technologies may bring to the unknown cause of the accident.Traditional safety analysis methods based on component fault models have limited ability to analyze complex human decisions,software errors,system accidents,and organizational risk factors.It is of great significance to improve civil aviation's operational safety and efficiency to solve complex systems' safety problems using the system thinking method of safety resilience.Therefore,this dissertation mainly researches Next Gen's safety resilience under complex operating environments.Building the Next Gen's safety resilience conceptual framework,this dissertation uses pre-analysis methods of safety resilience to in-depth research on the safety resilience functions' affordability and learning ability of air traffic control(ATC)system.Then introducing the natural language processing algorithm to realize the automatic classification of the ATC system's hazard,and uses the post-analysis methods of safety resilience to study the measurement problem of safety resilience function of the air transportation network's affordability,absorptive ability,and optimization ability under large-scale emergencies.The main research works of this dissertation are as follows:1.In view of the Next Gen safety resilience analysis method.Firstly,this dissertation mainly investigated the relevant definitions of resilience,resilience engineering,safety resilience assessment,and the systematic accident models.Secondly,based on the public safety triangle model and safety system shaping framework,and combined with the Next Gen characteristics,this dissertation proposed the Next Gen's safety resilience model,system safety resilience system conceptual model,and safety resilience assessment methods.Finally,the ATC system's safety resilience analysis framework was built.2.Aiming to identify and control new types of hazards in the safety resilience analysis framework of the ATC system.Firstly,this dissertation built different control seats Petri nets models in the operational control process of the ATC system and verified the simulation model using formal methods.Secondly,combined with the colored Petri net(CPN)model,the STPA risk analysis method is used to verify the case.By identifying system-level hazards and safety constraints,the state space reachable diagram is used to determine the reasons for unsafe control of the system,improving the STPA method's applicability.3.In view of the automatic classification of hazards in the safety resilience analysis framework of the ATC system.Firstly,a hazard classification index system of the ATC system based on the improved HFACS model is established.A professional dictionary in the field of safety of the ATC system is constructed.Secondly,a semi-supervised keyword extraction text classification algorithm based on improved TFIDF-Text Rank is proposed to solve fewer labels,fewer samples,and unbalanced samples of the ATC system's hazards text data.Then,this dissertation compared the model-based supervised text classification method and the proposed algorithm,and the experimental results show that the semi-supervised text classification way based on keywords is better than the model-based supervised approach for the ATC system.4.Aiming at resilience measurement of the air transportation network's safety resilience function under large-scale emergencies.Firstly,this dissertation constructed China's complex airport network and analyzed its essential structural characteristics.Based on the topology and data-driven,the dissertation proposed the resilience measurement method of airport network and constructed resilience metrics measurement model of the airport network.Finally,the resilience metrics of the critical nodes' strength of the airport network in China and Europe were analyzed and compared.The results show that the airport network recovery situation is significantly different from implementing different epidemic prevention policies,and the safety resilience capabilities are very different.