民航旅客订座规律分析及平行航班收益管理

发布时间：2018-03-20 01:12

本文选题：旅客规律　切入点：数据分析　出处：《南京航空航天大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着国内经济的转型升级,民航运输保持良好发展态势,新成立的航空运输企业众多,业务规模迅速扩张。民航旅客订座规律研究是航空运输企业对生产资源合理配置、市场销售策略有效运用的基础。在盈利能力较高航线,航空公司争相投入运力,使得一家公司在同一航线常有多个航班供旅客选择,因此,对平行航班收益管理方法的研究有较强的实际意义。本文主要研究民航旅客订座规律及平行航班收益管理。首先,对实际工作中的航线管理方法进行分析与总结,提出了通过客座率在订座期内的变化对旅客订座规律进行分析的方法。其次运用模型具体分析民航旅客订座规律,采用决策树方法,将旅客订座期按时间的由远及近划分为四期,并分析了各期旅客订座分布、旅客订座数的占比等特点;运用ARIMA模型对最终客座率建立预测模型,并观察旅客订座水平的平稳性、季节性、与历史数据的关系;建立不同订座期订座数据与最终客座率的关系模型,分析其对最终客座率的影响差异。最后,提出基于客座率预测的平行航班分期收益管理方法。将ARIMA模型根据历史规律得到的客座率作为期望客座率,以订座期数据的NARX神经网络预测结果作为对比值,结合各订座期旅客订座规律,提出平行航班收益管理方法。
[Abstract]:With the transformation and upgrading of domestic economy, civil aviation transportation keeps a good development trend, the newly established air transport enterprises are numerous and the business scale is expanding rapidly. The basis for the effective use of marketing strategies. On high-profitability routes, airlines compete for capacity, so that a company often has multiple flights on the same route for passengers to choose from. It is of great practical significance to study the methods of revenue management of parallel flights. This paper mainly studies the rules of airline passenger reservation and the revenue management of parallel flights. Firstly, it analyzes and summarizes the methods of airline management in practice. This paper puts forward a method to analyze the law of passenger reservation through the change of passenger rate during the reservation period. Secondly, the model is used to analyze the law of passenger reservation of civil aviation, and the decision tree method is adopted. The passenger reservation period is divided into four periods according to the time, and the characteristics of each period are analyzed, such as the distribution of passengers' reservation, the proportion of passengers' reservation number, and so on, and the prediction model of the final seating rate is established by using ARIMA model, and the stability of passenger reservation level is observed. The relationship between seasonality and historical data, the relationship between the data of different reservation periods and the final guest rate is established, and the influence of the data on the final guest rate is analyzed. This paper puts forward a method of managing the revenue of parallel flights by stages based on the prediction of passenger occupancy rate. The ARIMA model takes the passenger rate obtained by historical law as the expected passenger rate, and takes the forecast result of NARX neural network of the data of reservation period as the contrast value. Combined with the rules of passenger reservation in each reservation period, the paper puts forward the revenue management method of parallel flights.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F562

【参考文献】