Research On Methods Of Passenger Protection In Irregular Flight Based On Classification

With the rapid development of air transport business in our country, improving the service quality is an effective technology to enhance their own market competitiveness and build brand services. Due to the negative factors, such as bad weather, flight equipment failure, air traffic control, ground services ineffective and many others, irregular flights phenomenon is more and more frequent. If it often happens, the airlines would suffer huge economic losses, even worse, the passengers' trustworthiness to airlines would be severely threatened. Therefore, how to avoid the irregular flights risk to reduce costs, and propose timely emergency plan, has become a priority, and ensures the maintenance of the airline's reputation and customer base. This paper aims to study the itinerary recovery of passengers in irregular flight, based on the passenger classification with actual booking data. The main works are as follows:Analysis of the research background. Irregular flight will not only cause great inconvenience to passengers, but also become a major obstacle to improve service quality and reduce operational costs to airlines. Passengers flow recovery in irregular flight is raised to the question.Passenger classification. In this paper, it establishes a two-dimensional classification system for passengers. The first dimension classification system is for the passengers' behavior loyalty, and the second dimension applies the multivariate latent class model to study passenger classification with booking data PNR, based on the first dimension system.Recovery the disturbed passenger itinerary. This paper presents a recommendation model of passenger itinerary recovery, based on the passenger classification: disturbed passengers flow recovery model. First, systematically analyzes the cost caused by airline flight delays, the following is the constraints details of the passenger flow recovery model problem and model construct. Finally, propose a suitable model to solve the problem, as well as a method to determine the priority of travelers.The work of this paper is focused on the classification of airline passengers, and the recovery of disturbed passengers. Firstly, process the PNR attribute fields with variable cluster analysis. Then refine the concept of passenger behavior loyalty with attribute fields having high correlation, and give calculation methods. Thirdly, based on the data preprocessing of property fields having low correlation, introduce the construction and the results of passenger classification model in detail. Finally, based on the subdivision of flight delays cost, and combining the aircraft capacity constraints, passenger category and group bookings, construct passenger flow recovery model based on passenger category, and propose the suitable algorithm for this model.