| Safety is the important sign of social civilization and progress, also the foundation of civilaviation industry’s survival and sustainable development. In recent years, civil aviation industry ofChina has made certain achievements in regard to safety with the improvement of management andtechnique. The safety performance seems to be enlisted as one of the best throughout the world in theperspective of result. However, the overall safety foundation of China civil aviation is still relativeweak, safety management system is still under construction to achieve perfection, comprehensivesafety handling ability cannot match the rapid development of industry. In certain safety managementaspects, defects still do exist. There is a lot of work need to be done to catch up with developedaviation countries. According to related statistics, runway safety has been the most prominent issue ofChina civil aviation industry in the past10years. The research on runway safety is the inevitabledemands of the government and stakeholders when air traffic flow reaches to a certain extent. It isalso the only choice to break through the bottleneck in safety management.In view of the background, this paper studied the three key issues of runway safety, namelyrunway confusion, runway excursion and runway incursion under a unified framework with themethod of data driving. Aircraft accident analysis theory, failure diagnosis theory, the gray systemtheory, Bayesian network, artificial intelligence and so on were adopted comprehensively tosystematically and quantitatively research the causes for the above three issues. Accordingly, guidanceand strategy were raised to get the solution. Eventually a scientific and sound runway safetymanagement theory was formed. The result of this study will change our long-lasting runway safetymanagement mode which is mainly based on perceptual experience. It can offer theoretical guidanceto the systematical and targeted runway safety prevention measures for all the stakeholders such as theairlines, airports, air traffic management agencies and the authorities.The main contents are as follows:（1）Study on the event of aircraft’s mistaken alignment or landing on wrong runway. It dividedthe issue of runway confusion into two categories, i.e., departure from wrong runway and mistakenrunway alignment or landing on the wrong runway. Statistics showed that the latter one is theprominent problem of china civil aviation in recent years. SHEL model was adopted tosystematically analyze aircraft’s mistaken alignment or landing on wrong runway in the past12years. It was discovered that the model had significant deficiency. Therefore, SHEL model wasmodified into M-SHEL model, which was used to determine the causes of aircraft’s mistakenalignment or landing on wrong runway events. A sequence of all the factors was generated according to the level of influence based on gray relative analysis method. It was clear that the core link toprevent this kind of event from happening should focus on management factor in addition to crewand environment which could be identified easily.（2）Study on runway excursion for commercial flight. Data showed that the prominentappearances of runway excursion were runway overrun and runway veer-off after landing. Mechanicalanalysis was made toward take-off, landing and aborted take-off respectively, and the rolling distancewas deduced at the same time. It can be found in the result that aircraft overrun was related to thebuilt-in performance of a certain aircraft type, operation condition of flight and the control techniquesof the crew. A detailed analysis was followed based on the near two factors. As a result, a runwayoverrun prevention method was introduced. In order to improve human performance, a human factorintervention matrix was constructed according to the unsafe behaviors in runway overruns. Thismodel suggested that human’s unsafe behaviors require intervention from five levels, namelyorganization management, individual/team, technology, tasks and environment. Through evaluation offeasibility, acceptability, economy and effectiveness, the priority level of safety recommendations wasdetermined. In order to find out the key factors of runway excursion, it put forward a modelingmethod with the integration of fault tree and Bayesian network, and conversion algorithm betweenfault tree and Bayesian network. Compared with traditional fault tree analysis, the new method had astronger ability of modeling analysis, and it was closer to the actual situation of the events. It not onlymade full use of the advantage of logical analysis, but also realized the function of describingmulti-state event and probability calculation. It can improve the accuracy of quantitative analysis.Special attention should be given to the main risky factors such as standing water, reverse thrustbraking failure and complicated weather condition so as to cope with the issue of aircraft’s veer off.（3）Study on runway incursion. Based on the definition of runway incursion from internationalcivil aviation organization, the design frame on “the Protected Area of a Surface” was proposed. Thesurface protection of localizer, glide path station, the protection of runway strip and the runway endsafety area are needed to be considered in the planning of “the Protected Area of a Surface”. At thesame time, the linkage line between holding points at the aircraft taxiing side as well as a paved areawith the distance of75meters away from runway centerline on the non-taxiing side should also bepaid attention to. For the identification of hotspots, two quantitative study models of matrix methodand fuzzy clustering were put forward. Examples indicated that both of them have good applicabilityfor identification of hotspots. Combination of the two methods and taking their intersection canimprove the reliability of the results. Compared with the advanced aviation countries, comparativestudy showed that China has its own specific environment in terms of runway incursions. The main contributing factors of the runway incursions in China civil aviation industry can thus be illustrated asfollows: people or vehicle entering the runway without authorization due to poor management offlight area; the improper coordination between the civil and the military aviation agencies. Unsafeevent data have been studied quantitatively based on threat and error management model. The leastsquare was used to make multivariate regression analysis on the results and contributing factors. Withthe two element linear regression equation, the main factors of runway incursion can be identified. Acontrol model for runway incursion was introduced. It can be used as a management tool toconveniently grasp operation performance and make effective intervention at real time for those whoare in charge of managing.（4）Study on loss assessment of runway safety occurrence, incident investigation and situationprediction. In order to improve the understanding of the professionals, a new situation evaluationmethod was presented. It is more rational and objective to consider the observable and potentialdamages. Practice proved that the algorithm is convenient for operation and easy to master.Comparative analysis about case-based reasoning and rule-based reasoning has been made in order tofind their advantages and disadvantages, which established a kind of fusion reasoning method. Fordetermining the weight of each characteristic’s nature, analytic hierarchy process was improved byoptimal transfer matrix so as to satisfy the consistency requirement naturally without testing. Theintegrated techniques of rule-based reasoning and case-based reasoning can largely overcome theirshortcomings, which offered an intelligent-assisted decision making for runway safety investigation.Gray theory and BP neural network have been integrated to make runway safety prediction researchbased on Shapley weight distribution. The results show that the combined method has a certainsignificance to improve the accuracy of prediction and the situation of runway safety in china willremain serious in future. |
PDF Full Text Request