【作者】 刘刚；

【导师】 朱金福；

【作者基本信息】 南京航空航天大学 ， 交通运输规划与管理， 2009， 博士

【摘要】 随着我国经济的增长,航空运输业得到了飞速发展,近年来许多机场越来越繁忙并面临着越来越大的保障压力,国内一半以上的枢纽机场和干线机场的设计保障能力已进入饱和或接近饱和状态。在这种情况下,由于机场软件和硬件设施满足不了航空运输快速发展的需求,在机场发生的不安全事件和地面安全事故日益增多。诸多研究表明许多航空事故的发生都和地面保障因素有关,机场作为航空运输的主要地面活动场所,其安全保障对整个航空运输安全系统有着十分重要的意义。如何提高机场的安全保障能力,对机场运行保障中存在的安全风险进行识别、预测、评价、预警和对事故做好应急救援工作是机场管理当局非常关心的问题。本文根据系统工程理论将机场看作是一个复杂的巨系统,运用灰色系统理论和系统安全工程理论,综合研究各子系统安全运行保障的状况,对机场运行保障中存在的安全风险识别、预测、评价、预警和应急救援问题进行了研究。一、通过研究机场安全事故案例,在事故树定性分析的基础上,利用动态灰色关联方法分析机场机坪事故(安全违章事件)的人为失误因素。通过分析得出,影响机坪事故的人为失误主要原因为组织管理和个人原因,机场的机坪安全管理可根据此研究得出的结论有针对性地加以完善和改进。二、结合灰色GM(1,1)新陈代谢模型和马尔可夫链理论的优点,建立机场安全事故预测模型。灰色预测模型通过时序数据累加生成模块滤去原始序列中可能混入的随机分量,强化了新序列的规律性,新陈代谢后的数据信息更能反映当前系统的实际情况,再用马尔可夫链预测来弥补灰色预测的不足,提高了预测精度,为机场的安全预测增加了一种新的有效方法。三、结合中国民航现行的安全管理系统,建立了民航机场安全风险评价指标体系,通过安全评价量化分析民航机场安全风险。量化评价采用灰色多层次评价方法,实例分析结果表明,这是一种实用有效的风险评估方法。四、建立了机场安全风险预警模型。本文在灰色评价的基础上,建立了机场安全风险预警模型,模型将机场分成了十个子系统,机场的风险级别由十个保障子系统的最高安全风险预警级别决定。机场安全风险预警模型将对民航机场的安全状况起到温度计般的作用,作出简便、科学、准确的评价,具有较好的可操作性和实用性。五、运用鲁棒优化方法对机场应急救援设施建设选址问题进行了研究。如何在有限的时间内到达救援目的地,对保障航空飞行安全有着十分重要的意义。由于不确定因素的影响,确定和随机分析方法得到的解与实际的最优解往往偏差很大。鲁棒优化方法是解决不确定问题的有效方法,可使得到的解与各种可能情景下的最优解的最大偏差达到最小。更多还原

【Abstract】 With national economic increase, air transport industry developed rapidly. Recently, many airports are getting busier and facing heavier support pressure. The designed support capacity of more than half national hub airports and major airports has reached or approached saturated state. In this case, due to the airports’software and hardware facilities can’t meet the demand of air transportation’s fast development, insecurity incidents and ground security accidents happened in the airport increase day by day. A series of studies show that many aviation accidents are related to the ground guarantee factor. As the main ground activity place of air transportation, airport with its safety support plays a very important role in the whole air transportation security system. So, how to improve the safety assurance ability, to identify, to forecast, to estimate and to early warn the security risks existing in the airport operation and support, as well as the emergency rescue to the accidents, are the main problems concerned by the airport administering authority.Based on the system engineering theory of taking airport as a complicated giant system, this article comprehensively researches on the safety operation support status of all subsystems, using gray system theory and system safety engineering theory to study the identification, forecasting, estimation, early warning and emergency rescue of security risks existing in the airport operation and guarantee.First, we discussed many airport security accident cases. Then based on qualitative analysis of accident tree, we analysed the human error factors of apron accidents or insecurity incidents by the method of dynamic gray correlation. Through that, we find that the main causations of human error are organization management and personal reasons, according to which the security management of apron can be improved and consummated pertinently.Second, combined the advantages of gray GM (1,1) metabolic model and Markov chain theory, we set up a airport security accident forecast model. Through the module built by the time series data accumulation, the gray forecast model sieves the stochastic components of original data series, strengthens the orderliness of new sequence and uses Markov chain to offset the shortage of gray forecast, to increase the prediction accuracy. So that we proposed a new effective method for airport security forecast which can reflect the actual situation of current system.Third, combined with active security management system of CAAC, we establish an evaluation index of civil airport security risk, by which we can quantitatively analyse civil airport security risk, and adopt grey multi-level evaluation method to evaluate the security risk. The result shows that it is a practical and effective method of risk evaluation.Fourth, on the basis of gray evaluation, we set up an airport security risk early warning model, which divides the airport into ten subsystems, and the airport’s risk level is determined by the highest security risk alarm level of the support subsystems. This model is operable and practical, which will figure as a thermometer for the safety situation of civil airports, evaluating easily, scientifically and exactly,Finally, we studied and built the robust optimization method of support facility location problems about airport emergency rescue. How to arrive at the succor destination in the limited time is very important for the aviation flight security. Due to the effect of uncertain factors, the results of deterministic and stochastic method usually are far from the actual optimal one. Robust optimization method is an effective way to solve the uncertain problems, the maximum deviation between the result of which and the optimal result in all kinds of possible scenarios can be least.更多还原