【作者】 王巡；

【导师】 蒲云；

【作者基本信息】 西南交通大学 ， 物流工程， 2013， 博士

【摘要】 航空物流运输是我国目前最有发展前景的一种物流运输方式。因为航空运输快速且价格高的特点,随着我国经济的快速发展,人民的生活水平有了长足进步,对于出行方式的选择,也从过去局限于陆路交通发展为越来越多的倾向于对航空交通方式的选择。航空交通的发展,不仅是国家经济发展和人民生活水平提高的反应,更是从侧面反应了一个国家科技水平和军事实力。因此,民航运输业的兴衰,从某种意义上关系到国家的整体发展和现代化进程。所以对于航空业和航空公司航班管理的研究就显得更加重要。论文首先论证了航空旅客消费行为分析在当今航空公司航空物流管理研究的重要性,其次基于前人的基础,本论文创新性利用数据手段(弹性计算以及聚类分析)分析影响旅客选择航班的因素,并将旅客消费订单进行细分,对于不同消费者提出不同的营销意见,有助于航空公司的市场营销。同时,通过消费者订单数据(非问卷调查)建立消费者流失模型,对不同消费者进行预测,有助于公司及时采取营销手段,预防消费者流失。然后,本论文研究了机票的超售问题。首先介绍了航空公司进行机票超售的原因和可能的影响因素；其次通过对机票超售策略的主要决定因素No-Show率或售乘率的分析,解释了现有的机票超售策略中存在的现象和问题,本论文还对No-Show率和售乘率进行了预测,使得航空公司可以以此进行机票超售管理,以实现航空公司利润最大化的目的。本论文还研究了航班延误的波及效应。航班延误在社会上越来越受到人们的关注,传统的文献已经对其进行了一系列的研究。与传统文献不同的是,本论文研究了航班延误的波及效应,即一架航班的延误对于其他航班的延误可能性的影响,本论文称之为波及效应。更加具体地,本论文进行了如下几个假设的验证,1)上一站的延误航班越多,飞机晚到造成的延误航班次数越多；2)首次延误航班的延误时间越长,航班延误波及效应越显著；3)机场自身对航班正点率有影响。最后,本论文还重点研究了市场旅客流量预测问题和航班频率优化问题。首先以航空市场旅客流量的基本特征为出发点,建立季节性组合预测模型,并以此预测相关航线上的旅客流量,然后以成本最小化为目标,建立航班频率优化模型。在此基础上,选用中国南方航空股份有限公司大连分公司的相关航线数据为例进行实证分析。本论文的创新点在于进一步优化航空市场旅客流量的预测模型和航班频率优化模型,并且从实证角度加以证明其可行性。本论文在以下几个方面对相关领域的文献做出了贡献。首先,对于航空公司以消费者为基础的研究文献几乎没有以消费者数据和航班数据为资料,而本论文创新性地提出从消费者消费行为数据、航班相关数据等方面进行研究,一方面,相较于理论型或者逻辑分析型的文章,本论文具有更加真实的实证证据。另一方面,消费者行为数据和航班相关数据是消费者行为和航空公司航班管理的真实数据反应,从其中所总结出来的现象就是可以代表消费者行为和航空公司航班管理的行为。其次,基于本论文数据的优势,本论文得以在实证分析部分加入聚类分析、季度趋势等先进的实证分析方法,为本论文的分析和结论添加更加具有说服力的证据。再次,本论文对航空公司航班管理的出发点相较于传统领域的文献具有重大创新,即以消费者行为分析为出发点,这样进行分析的好处就是,消费者是航空公司的主要服务对象,消费者的行为趋势对航空公司的政策都有直接的影响,所以本论文从此出发,更能体现航空公司本身的航班管理目的。更多还原

【Abstract】 Air transport has been and will be one of the most important transportation means in our country. Along with air transport’s two main characteristics, high speed and high cost, people in China nowadays have chosen air transport as their main travel solution much more frequently than just two decades ago, partially because of the improvement in people’s living standard induced by this country’s rapid economical growth. On the one hand, considerable growth in air transportation indicates our improving living standard or increasing income; on the other hand, it’s also a piece of significant evidence in the strength or this country’s technology and military power. Thus, development of our air transportation industry is strongly related to the process of this country’s long-term development and modernization. In this sense, research on this topic is and will be necessary.In our paper, we first show why consumer behavior analysis is important and necessary to the airplane company governance research. Based on existing literature on this topic, we innovatively utilize the data analysis methods (elasticity calculation and clustering analysis) to show the determinants of consumer choices. We carefully group our consumer order data to separately make policy implication and suggestion for difference consumer group, which is essential to air company decision making. Meantime, we establish a model to describe and forecast the loss of consumers (not based on questionnaire summary). We predict the probability of losing a specific group of consumer, and this will be helpful for an airplane company to make effective marketing policy and prevent losing more costumers. Then we test the so-called the method of ticket over selling. We first illustrate why the airplane company will use such an approach and what determines this choice. We’ve explained the determinants of the no-show rate and more patterns in the undergoing practice of tickets over-selling. We also make no-show rate forecasting, in order to make a more effective over-selling strategy. We also looked at the effects of flight delay spreading. Flights delays have attracted more and more attention from society, and we also have some conventional literature on this topic. Different from those, we researched the effects of flights delay, i.e. the effects of one flight delay on other flights. More specifically, we’ve made the following hypothesis to test:1) more flights in this station will be delayed if more are delayed in the last station;2) if the time length for the first delay is longer, the spread effects are stronger;3) airport itself does impact on flights on-time rate. In the end, we additionally looked on the passengers flow forecasting and flights frequency optimization problem. We first establish a seasonal forecast model based on the passengers flow basic characteristics, then we predict the passenger load for each air line, then the frequency optimization model is established for profit maximization.Our research has made contributions to the literature in the following aspects. First, compared to the fact that few existing literature has access to the customer consumption and order data, we innovatively start our research using these recourses. On the one hand, compared to the logistics and theoretic analysis, we provide more realistic evidence for our argument; on the other hand, consumption and order data is the true reflection of consumer behavior, which is a direct evidence for consumer behavior analysis. Second, based on our data source advantage, we utilized more modern empirical methodology, including clustering analysis, seasonal analysis etc. Third, we contributed to the literature in the sense that we make our starting point from the consumer behavior analysis because customers are the focus of the airplane companies and we can stand on the viewpoint of them, i.e. profit maximization to solve the problem.更多还原