Research On Airline Revenue Management Based On Data-driven Kaplan-meier Method

With the increase of residents' spending ability and the motivation of people's travel demand,the demand for China's air passenger transport market continues to increase,bringing new opportunities for development.But the new opportunities always faced with more intense market competition in this industry.in order to be in the advantage of fierce market competition Domestic and international airlines basically use revenue management strategies to raise their income levels.However,when domestic and foreign airlines use revenue management strategies,they mostly assume that customer demand obeys the fixed random distribution,assuming that there will be a certain deviation between forecasted demand and real demand,This deviation will bring great problems to airline revenue management.This is the problem that the article wants to solve.In addition,the paper mainly focuses on the problem of air passenger space control revenue management,and further studies the simple oversale problem considering the risk of rejection on the basis of the research of cabin control.First of all,this paper reviews the current domestic and foreign research on seat inventory control and overbooking,and introduce the data-driven Kaplan-Meier method to study the problem of assuming that the customer demand obeys the fixed random distribution and the real de mand error.In the actual process,an explicit function containing censored data information is introduced to make the selected historical demand data more in line with the actual demand,so that the existing flight capacity resources can be effectively used to allocate the class inventory to the corresponding customer market at a reasonable price.It is hoped that by using Kaplan-Meier method to reduce the impact of the actual uncertainty of customer demand on the revenue generated by air passenger transport.Secondly,in order to make the research more closed to actual situation of aviation passenger transport,on the basis of the application of Kaplan-Meier method to the study of two price class space control problems,and the problem of cabin control of multiple price classes is further studied.Because the seat inventory control is closely related to the problem of overbooking,the Kaplan-Meier method is also used in this paper to study the overbooking problem of a single flight segment on a single flight.Finally,some numerical simulations on the allocation and overbooking problems based on Kaplan-Meier method are carried out? The numerical analysis section of the two price class control sections compares the results obtained using the Kaplan-Meier method with the numerical results obtained from the Littlewood rule and the unilateral pruning data drive method.The numerical analysis section of the multi-price class space control section compares the results obtained using the Kaplan-Meier method with the numerical results obtained by the EMSR-a heuristic algorithm and the single-sided pruning data drive method.The numerical analysis section of the simple overbook ing problem that takes into account the risk of rejection,compares the results obtained using the Kaplan-Meier method with the results obtained from the binomial distribution and the numerical results obtained from the unilateral pruning data-driven method.the analysis results show that it is feasible to apply Kaplan-Meier method to the problem of cabin allocation and overbooking,because applying the Kaplan-Meier method to solve the aviation revenue management problem can not only quickly add and deal with the newly generated data in the current period,but also the model does not need to be modified,The calculation process is quicker and easier.This method can help air passenger managers to rapidly realize the strategy of seat inventory control and overbooking In the background of the big data age,thus achieve the goal of maximizing the expected profit of air passenger transportation.