An air ticket sales trend deep learning analysis system

Abstract

The application relates to the technical field of air ticket sales analysis, and discloses an air ticket sales trend deep learning analysis system.The specific implementation steps of the application include the following steps: collecting and cleaning historical air ticket sales data, operation data and passenger data of an airline through an air ticket related data collection module; an air ticket related data processing module analyzes key air ticket attributes and operation attributes of different passenger cabin grades, calculates passenger cabin probability indexes, and forms historical sales data sets of different passenger cabin grades; an air ticket sales trend analysis module constructs a sales volume prediction model and optimizes the model, so as to predict future sales volumes of different passenger cabin grades; and an air ticket sales strategy regulation and control module determines an optimal sales strategy by using a reinforcement learning algorithm, combining an economic state space, a sales strategy action space and a sales strategy value function.The application can effectively predict future sales trends, optimize sales strategies and improve the overall operation efficiency of an airline.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of air ticket sales analysis, in particular to an air ticket sales trend deep learning analysis system. BACKGROUND

[0002] With the rapid development of the global aviation market, air ticket sales have gradually shifted from traditional manual booking to an online sales model centered on the Internet. The rapid development of Internet technology and the popularity of mobile payments enable airlines to quickly and conveniently realize air ticket sales through their own platforms or third-party platforms, improving the interaction efficiency between airlines and consumers. In order to enable airlines to better understand passenger demand, optimize route layout and ticket pricing strategies, air ticket sales trend analysis technology has emerged. This technology can accurately predict future sales trends by analyzing historical sales data, market demand fluctuations, and consumer purchasing behavior, thereby optimizing pricing strategies, adjusting flight schedules, and improving passenger satisfaction.

[0003] Currently, existing sales trend analysis technology focuses on analyzing the impact of holidays, seasonal fluctuations, and passenger historical travel behavior on air ticket sales trends, ignoring the impact of specific flight schedules, airline service quality, and passenger cabin type selection and flight purposes on air ticket sales, which reduces the accuracy of air ticket sales trend prediction and may lead to subsequent flight regulation and air ticket pricing errors, further causing customer loss and economic losses.

[0004] Therefore, an air ticket sales trend deep learning analysis system is proposed. SUMMARY

[0005] The purpose of the present application is to provide an air ticket sales trend deep learning analysis system that collects and cleans historical air ticket sales data, operation data, and passenger data of airlines through an air ticket related data acquisition module; then, an air ticket related data processing module analyzes key air ticket attributes and operation attributes of different cabin grades and calculates passenger seat probability indexes to form historical sales data sets of different cabin grades; then, an air ticket sales trend analysis module builds a sales volume prediction model and optimizes it to predict future sales volume of different cabin grades; finally, an air ticket sales strategy regulation module determines the optimal sales strategy using reinforcement learning algorithms combined with economic state space, sales strategy action space, and sales strategy value functions. The present application can effectively predict future sales trends, optimize sales strategies, and improve the overall operational efficiency of airlines.

[0006] An air ticket sales trend deep learning analysis system comprises:

[0007] The ticket-related data collection module collects the airline's historical ticket sales data, operation data, and passenger data, and performs data labeling and data cleaning to obtain historical ticket sales data, historical operation data, and historical passenger data;

[0008] The air ticket-related data processing module obtains the air ticket attribute correlation based on the air ticket attributes of the historical air ticket sales data and the historical sales data of different cabin classes, and determines the key air ticket attributes of different cabin classes by combining the sales volume-air ticket attribute regression model; determines the key operational attributes of different cabin classes by building a sales volume-operation attribute regression model and performing operational attribute screening; calculates the passenger class probability index for different passengers, divides the historical passenger data of different cabin classes, and combines the data corresponding to the key air ticket attributes and key operational attributes to obtain the historical sales data set of different cabin classes;

[0009] The ticket sales trend analysis module builds a sales volume forecasting model, trains and optimizes the model on historical sales data sets of different cabin classes, obtains the trained sales volume forecasting model, and predicts future sales volumes of different cabin classes;

[0010] The air ticket sales strategy control module sets the economic state space, sales strategy action space and sales strategy value function, and adjusts the sales strategy based on the economic state space; calculates the adjusted air ticket profit index, updates the sales strategy value function through the reinforcement learning algorithm, and determines the optimal sales strategy.

[0011] Preferably, the ticket sales data includes the take-off time, landing time, flight duration, date, ticket fare and sales volume of each flight; the operating data includes aircraft type, failure rate and cabin service quality; passenger data includes cabin class, ticket purchase frequency and ticket purchase purpose; the data marking process is to mark the corresponding holidays according to the date of each flight, and replace the date attributes in the ticket sales data with holidays.

[0012] Preferably, the specific implementation process of the ticket-related data processing module includes:

[0013] Dividing the sales volume in the historical ticket sales data according to cabin class to obtain historical sales volume of economy class, historical sales volume of business class, and historical sales volume of first class;

[0014] For the historical ticket sales data, the ticket attribute correlation between different ticket attributes is calculated using the Pearson correlation coefficient; the sales volume-ticket attribute regression model is constructed, and the degree of influence of different ticket attributes on the historical sales volume of different cabin classes is determined by fitting the historical data, thereby obtaining the influence of economy class ticket attributes, business class ticket attributes, and first class ticket attributes; based on the ticket attribute correlation, the ticket attributes corresponding to different cabin classes are screened to determine the key ticket attributes for economy class, business class, and first class.

[0015] For the historical operating data, construct the sales volume-operation attribute regression model, and determine the key operating attributes of economy class, business class, and first class by fitting the historical data and screening the operation attributes;

[0016] For the historical passenger data, the passenger class probability index is calculated based on cabin class, ticket purchase frequency, and ticket purchase purpose. The specific calculation formula is:

[0017]

[0018] Among them, PC i represents the passenger class probability index of passenger i; Indicates cabin class sequence C i The value of cabin class c when it appears the most times, f() is the frequency function; exp() represents the exponential function with a natural constant as the base; β represents the ticket purchase frequency sequence F i Average frequency of ticket purchases The exponential weight of the ticket; p represents the ticket purchase purpose sequence P i The specific code value of the ticket purchase purpose; α1 represents the cabin class sequence C of passenger i i The impact factor of α2 represents the ticket purchase frequency sequence F of passenger i. i The impact factor of α3 represents the ticket purchase purpose sequence P of passenger i. i The impact factor is 0<α2<α3<α1<1 and α1+α2+α3=1;

[0019] dividing the historical passenger data into economy class passenger historical data, business class passenger historical data, and first class passenger historical data according to the passenger class probability index;

[0020] The passenger historical data, key ticket attributes and key operation attribute data corresponding to different cabin classes are integrated to obtain the economy class historical sales dataset, business class historical sales dataset and first class historical sales dataset.

[0021] Preferably, the sales volume prediction model includes a multi-dimensional feature extraction layer, a feature analysis and fusion layer, and a sales trend prediction layer;

[0022] The multidimensional feature extraction layer is used to extract ticket sales features, flight operation features, aircraft operation features and passenger features from the historical sales dataset of different cabin classes;

[0023] The feature analysis and fusion layer is used to analyze the impact of the flight operation characteristics, the aircraft operation characteristics, and the passenger characteristics on the ticket sales characteristics of different cabin classes, and perform feature weighted fusion;

[0024] The sales trend prediction layer is used to predict the future sales volume of different cabin classes.

[0025] Preferably, the specific implementation process of performing feature analysis and feature weighted fusion through the feature analysis and fusion layer includes:

[0026] The flight operation characteristics and the passenger characteristics are input into the cross-attention mechanism to capture the potential impact of different flights on passengers' ticket purchasing behavior and obtain the passenger-flight preference weight; the aircraft operation characteristics and the passenger characteristics are input into the cross-attention mechanism to analyze the preferences of different passengers for aircraft operation conditions and obtain the passenger-aircraft preference weight; the passenger characteristics, the aircraft operation characteristics and the flight operation characteristics are weightedly fused in combination with the passenger-flight preference weight and the passenger-aircraft preference weight to obtain the passenger preference characteristics; through the LSTM network, the influence of the passenger preference characteristics on the ticket sales characteristics is explored to obtain the multi-dimensional fusion characteristics of ticket sales.

[0027] Preferably, when adjusting the sales strategy in the sales strategy action space, the future sales volume of different cabin classes is re-predicted by the ticket sales trend analysis module to update the economic status space;

[0028] The economic status space includes future sales volume of economy class, future sales volume of business class, future sales volume of first class, oil prices and flight distance; the sales strategy action space includes ticket fare adjustment, flight schedule adjustment and aircraft type adjustment.

[0029] Preferably, the specific implementation process of calculating the adjusted ticket profit index, updating the sales strategy value function, and determining the optimal sales strategy includes:

[0030] After adjusting the sales strategy, the ticket profit index is calculated; the sales strategy value function is updated through the reinforcement learning algorithm to obtain the optimal sales strategy; the specific formula for updating the sales strategy value function is:

[0031]

[0032] R u =es u ·ep u +bs u bp u +fs u ·fp u -(es yc ·ep ys +bs yc bp ys +fs yc ·fp ys )+l u Δop;

[0033] Among them, Q(S u ,A u ) represents the current economic state S u Next, select Sales Strategy Action A u The value of the sales strategy is, Q() is the sales strategy value function; γ represents the learning rate and 0<γ<1; R u Represents the current economic state S u Next, select Sales Strategy Action A u The instant reward is the ticket profit index; δ represents the discount factor and 0<δ<1; Indicates the next economic state S u+1 The maximum value of the next value, Indicates the next economic state S u+1 Sales strategy action when the next value is the largest; u ,bs u and fs u Respectively represent the selection of sales strategy action A u The future sales volume of economy class, business class and first class at the time of ep u , bp u and fp u Respectively represent the selection of sales strategy action A u Economy class airfare, business class airfare and first class airfare at that time; yc ,bs yc and fs yc They represent the future sales volume of economy class, business class and first class when no sales strategy adjustment is made; ep ys , bp ys and fp ys Economy class airfare, business class airfare and first class airfare before sales strategy adjustment; u Represents the current economic state S uflight distance under the current economic state S u the difference between the oil price under the current economic state S and the oil price under the last economic state.

[0034] Compared with the prior art, the application has the following beneficial effects:

[0035] 1、The flight ticket related data processing module improves the accuracy of flight ticket sales trend prediction by combining the sales-volume-flight ticket attribute regression model, the sales-volume-operation attribute regression model, and the passenger cabin probability index. Through detailed historical data analysis of different passenger cabin grades (economy class, business class, and first class), key flight ticket attributes and operation attributes are identified, and accurate division and prediction are performed according to the passenger's ticket buying frequency, destination, and passenger cabin grade influencing factors. The module can accurately reflect the sales trend of each passenger cabin grade under different conditions, improving the prediction ability of flight ticket sales data. Through the integration of the data set, more reliable data support can be provided for subsequent flight control and flight ticket pricing.

[0036] 2、The sales volume prediction model can effectively improve the accuracy of flight ticket sales trend prediction through multi-dimensional feature analysis and feature weighted fusion. By using the cross-attention mechanism to analyze and fuse different flight operation features, aircraft operation features, and passenger features, the potential rules of passenger behavior and preferences can be deeply mined, and the mutual relationship between different flights and passengers can be accurately captured; by combining the LSTM network to further analyze the deep influence of passenger preferences on flight ticket sales, multi-dimensional information fusion and trend prediction are realized, and the accuracy of sales volume prediction is improved.

[0037] 3、The application optimizes the sales strategy by combining the economic state space, the sales strategy action space, and the sales strategy value function, and combining the reinforcement learning algorithm, thereby effectively adjusting the subsequent flight and flight ticket pricing. Based on the accurate prediction of future sales volume by the flight ticket sales trend analysis module, the economic state space can be updated in real time to ensure that the sales volume of different passenger cabin grades is consistent with market demand, avoiding errors caused by market fluctuations. By optimizing the sales strategy, the flight ticket profit index is maximized, which not only improves the operation efficiency and revenue of the flight, but also effectively reduces the risk of customer loss and economic loss. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 A system structure diagram of a flight ticket sales trend deep learning analysis system provided for the embodiments of the application;

[0039] Figure 2 A step flowchart of a flight ticket sales trend deep learning analysis system provided for the embodiments of the application;

[0040] Figure 3A schematic diagram of the structure of the sales volume forecasting model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0042] To effectively manage and adjust its ticket sales strategy, Airline A introduced a deep learning analysis system for ticket sales trends. This system analyzes ticket sales trends for different cabin classes. To illustrate the effectiveness of the proposed deep learning analysis system, we will provide a detailed description using the accompanying figures and the following two examples.

[0043] Example 1

[0044] The embodiment of the present application discloses a deep learning analysis system for ticket sales trends, which is used to analyze the ticket sales trends of airline A and thus optimize sales strategies. Figure 1 and Figure 2 , which are respectively a system structure diagram and step flow chart of a deep learning analysis system for air ticket sales trends; the specific implementation steps of the system include: S1, collecting air ticket sales data, operation data and passenger data in historical time, and performing data annotation and data cleaning; S2, constructing a sales volume-air ticket attribute regression model and a sales volume-operation attribute regression model to determine the key air ticket attributes and key operation attributes of different cabin classes; dividing the passenger historical data through the passenger class probability index to obtain the historical sales data set of different cabin classes; S3, constructing a sales volume prediction model and predicting the future sales volume of different cabin classes; S4, setting the economic state space, sales strategy action space and sales strategy value function, and adjusting the sales strategy in combination with the economic state space; S5, calculating the adjusted air ticket profit index, optimizing the sales strategy through the reinforcement learning algorithm, and determining the optimal sales strategy.

[0045] Furthermore, the ticket-related data collection module collects the ticket sales data, operation data and passenger data of Airline A in the historical period, and performs data annotation and data cleaning to obtain historical ticket sales data, historical operation data and historical passenger data, corresponding to the above-mentioned S1 step; ticket sales data includes the take-off time, landing time, flight duration, date, ticket price and sales volume of each flight; operation data includes aircraft type, failure rate and cabin service quality; passenger data includes cabin class, ticket purchase frequency and ticket purchase purpose; the data annotation process is to mark the corresponding holidays according to the date of each flight, and replace the date attributes in the ticket sales data with holidays.

[0046] Tables 1, 2, and 3 list some examples of economy class ticket sales data, operational data, and passenger data.

[0047] Table 1. Some examples of economy class ticket sales data

[0048]

[0049] Table 2. Some examples of operational data

[0050] Flight number Aircraft type Failure rate (%) Cabin service quality AF2288 Airbus 320 0.05 8.5 AF2264 Boeing 737-800 0.02 9 AF2248 Airbus 321 0.1 7.5 AF2276 Boeing 737 0.03 8

[0051] Table 3. Some examples of passenger data

[0052] Flight number Cabin class Ticket purchase frequency (times / year) Purpose of ticket purchase AF2288 Economy Class 10 business trip AF2264 Business Class 5 travel AF2248 Business Class 8 business trip AF2276 Economy Class 6 visiting relatives

[0053] By collecting and processing Airline A's historical ticket sales, operational, and passenger data, a solid data foundation can be established for subsequent demand forecasting and optimization decisions. The data cleaning and annotation process not only eliminates noise from the raw data but also introduces a new dimension to the ticket sales data by annotating holidays. Combining ticket sales, operational data, and passenger behavior data reveals underlying flight patterns, such as the impact of high-demand periods, aircraft compatibility, and service quality on passenger selection, providing data support for subsequent ticket sales analysis.

[0054] Furthermore, corresponding to the above step S2, the specific implementation process of the ticket-related data processing module includes:

[0055] The sales volume in the historical ticket sales data is divided according to the cabin class to obtain the historical sales volume of economy class, business class and first class;

[0056] For historical ticket sales data, the Pearson correlation coefficient was used to calculate the correlation between different ticket attributes. A sales volume-ticket attribute regression model was constructed. By fitting historical data, the influence of different ticket attributes on the historical sales volume of different cabin classes was determined, and the influence of economy class ticket attributes, business class ticket attributes, and first class ticket attributes were obtained. The specific formula of the sales volume-ticket attribute regression model is:

[0057] AS j =κ j,1 ·T dep +κ j,2 ·T land +κ j,3 ·D flig +κ j,4 H+κ j,5 AT j +ε j ;

[0058] Among them, AS j represents the ticket sales volume of cabin class j, where cabin class j is economy class, business class, or first class; T dep represents the departure time of the flight; j,1 T represents the influence factor of departure time on the ticket sales volume of cabin class j, that is, the influence of departure time on cabin class j ticket attributes; land represents the landing time of the flight; κ j,2 D represents the impact factor of landing time on ticket sales of cabin class j; flig represents the flight duration; κ j,3 represents the impact factor of flight duration on ticket sales of cabin class j; H represents the holiday corresponding to the flight departure date; κ j,4 AT represents the impact factor of holidays on ticket sales of cabin class j; j represents the airfare of cabin class j in flight; κ j,5 represents the impact factor of air ticket prices on the ticket sales volume of cabin class j; ε j The residual term representing the influence of the ticket attribute of cabin class j;

[0059] Based on the ticket attribute correlation, the ticket attributes corresponding to different cabin classes are screened to determine the key ticket attributes for economy class, business class, and first class. If the ticket attribute correlation of two ticket attributes is greater than a predetermined correlation threshold, the ticket attribute influences of the two ticket attributes are compared, and the ticket attribute with the smaller ticket attribute influence is deleted.

[0060] For historical operating data, a sales volume-operation attribute regression model was constructed. Through historical data fitting and operation attribute screening, the key operating attributes of economy class, business class, and first class were determined. The specific formula of the sales volume-operation attribute regression model is:

[0061] AS j =ε j,1 LX airc +ε j,2 ·FR airc +ε j,3 CS+η j ;

[0062] Among them, LX airc Indicates the aircraft type of the flight; j,1 The factor that affects the sales volume of tickets of cabin class j is represented by aircraft type; FR airc represents the failure rate of the aircraft corresponding to the flight; ε j,2 represents the impact factor of the failure rate on the ticket sales of cabin class j; CS represents the cabin service quality of the corresponding aircraft of the flight; ε j,3 The factor that influences cabin service quality on ticket sales of cabin class j; η j The residual term represents the influence of operating attributes on the sales volume of tickets for cabin class j;

[0063] The operation attribute screening operation is to analyze the impact factor of the operation attribute on the ticket sales volume of cabin class j; if the corresponding impact factor is less than the predetermined impact threshold, the operation attribute is deleted;

[0064] For historical passenger data, the passenger class probability index is calculated based on cabin class, ticket purchase frequency, and ticket purchase purpose. The specific calculation formula is:

[0065]

[0066] Among them, PC i represents the passenger class probability index of passenger i; Indicates cabin class sequence C i The value of cabin class c when it appears the most times, f() is the frequency function; exp() represents the exponential function with a natural constant as the base; β represents the ticket purchase frequency sequence F i Average frequency of ticket purchases The exponential weight is set according to expert experience; p represents the ticket purchase purpose sequence P i The specific code value of the ticket purchase purpose; α1 represents the cabin class sequence C of passenger i i The impact factor of α2 represents the ticket purchase frequency sequence F of passenger i. iThe impact factor of α3 represents the ticket purchase purpose sequence P of passenger i. i The impact factor is 0<α2<α3<α1<1 and α1+α2+α3=1;

[0067] According to the passenger class probability index, the historical passenger data is divided into economy class passenger historical data, business class passenger historical data, and first class passenger historical data; specifically, if the passenger class probability index belongs to the economy class coding range, the historical passenger data is divided into economy class passenger historical data; if the passenger class probability index belongs to the business class coding range, the historical passenger data is divided into business class passenger historical data; if the passenger class probability index belongs to the first class coding range, the historical passenger data is divided into first class passenger historical data;

[0068] The passenger historical data, key ticket attributes and key operation attribute data corresponding to different cabin classes are integrated to obtain the economy class historical sales dataset, business class historical sales dataset and first class historical sales dataset.

[0069] By granularly segmenting historical ticket sales data and integrating it with regression models based on ticket and operational attributes, we can accurately analyze the impact of various factors on sales. Utilizing the passenger class probability index, we can segment passenger groups and gain a deeper understanding of passenger demand across different cabin classes. This process not only improves the reliability of subsequent sales forecasts but also enhances airlines' responsiveness to market dynamics.

[0070] Furthermore, corresponding Figure 2 In step S3, the ticket sales trend analysis module builds a sales volume forecasting model, trains and optimizes the model on historical sales datasets of different cabin classes, obtains a trained sales volume forecasting model, and predicts future sales volumes of different cabin classes.

[0071] The sales volume forecasting model includes a multi-dimensional feature extraction layer, a feature analysis and fusion layer, and a sales trend forecasting layer. Figure 3 , which is a structural diagram of the sales volume forecasting model;

[0072] Multidimensional feature extraction layer, used to extract ticket sales features, flight operation features, aircraft operation features, and passenger features from historical sales datasets of different cabin classes;

[0073] The feature analysis and fusion layer is used to analyze the impact of flight operation characteristics, aircraft operation characteristics, and passenger characteristics on ticket sales characteristics of different cabin classes, and perform weighted fusion of features;

[0074] The sales trend forecasting layer is used to predict future sales of different cabin classes.

[0075] Specifically, the multi-dimensional feature extraction layer uses a convolutional neural network to extract different features from the historical sales data of different cabin classes. The sales trend prediction layer captures the temporal information of the output features of the feature analysis fusion layer through the RNN network. Based on the fully connected layer and the prediction head, the future sales volume of different cabin classes is predicted. The predicted result of future sales volume can be the sales volume of a certain day in the future or the sales volume of each day in a certain time period in the future.

[0076] In order to clearly illustrate the accuracy of the sales volume forecasting model proposed in the present invention, referring to Table 4, the economy class sales volume forecast results for different dates are exemplarily given.

[0077] Table 4. Some examples of economy class sales forecast results for different dates

[0078] date Forecast sales volume (pieces) Actual sales volume (pieces) Forecast error rate (%) 2024 / 1 / 15 150 145 3.45 2024 / 2 / 15 180 175 2.86 2024 / 3 / 15 130 120 8.33 2024 / 4 / 15 160 155 3.23 2024 / 5 / 15 146 153 4.58

[0079] By building a sales volume forecasting model and training and optimizing it on historical sales datasets for different cabin classes, we can accurately predict future sales volumes for each cabin class. The multidimensional feature extraction layer effectively extracts ticket sales characteristics, flight operation characteristics, aircraft operation characteristics, and passenger characteristics, thereby comprehensively considering the impact of multiple factors on sales. The feature analysis and fusion layer performs weighted fusion of different features, ensuring that each feature receives a reasonable weight and influence in the forecasting process, making the model more accurate and adaptable. The sales trend forecasting layer can accurately predict sales trends based on the optimized model, providing a reliable decision-making basis for subsequent sales strategy adjustments, thereby improving sales efficiency and reducing seat vacancy rates.

[0080] Furthermore, the specific implementation process of feature analysis and feature weighted fusion through the feature analysis fusion layer includes:

[0081] Flight operation characteristics and passenger characteristics are input into the cross-attention mechanism to capture the potential impact of different flights on passengers' ticket purchasing behavior and obtain passenger-flight preference weights; aircraft operation characteristics and passenger characteristics are input into the cross-attention mechanism to analyze different passengers' preferences for aircraft operation conditions and obtain passenger-aircraft preference weights; passenger characteristics, aircraft operation characteristics and flight operation characteristics are weightedly fused by combining passenger-flight preference weights and passenger-aircraft preference weights to obtain passenger preference characteristics; through the LSTM network, the impact of passenger preference characteristics on ticket sales characteristics is explored to obtain multi-dimensional fusion characteristics of ticket sales.

[0082] By combining a cross-attention mechanism with an LSTM network, this embodiment of the application can more comprehensively capture passengers' personalized needs and preferences, facilitating in-depth analysis of the potential impact of different flight and aircraft operating conditions on passenger purchasing behavior. By weightedly integrating passenger, flight, and aircraft operating characteristics, it can better describe passengers' multi-dimensional preferences and improve prediction accuracy.

[0083] Furthermore, the ticket sales strategy control module sets the economic state space, sales strategy action space and sales strategy value function, and adjusts the sales strategy in combination with the economic state space, corresponding to Figure 2 Step S4; when adjusting the sales strategy in the sales strategy action space, the future sales volume of different cabin classes is re-forecasted through the ticket sales trend analysis module to update the economic state space;

[0084] Among them, the economic status space includes future sales volume of economy class, future sales volume of business class, future sales volume of first class, oil prices and flight distance; the sales strategy action space includes air ticket fare adjustments, flight schedule adjustments and aircraft type adjustments.

[0085] Specifically, the agent first observes the current economic state, namely, factors such as future sales volume for different classes and fuel prices. Next, based on the current state, the agent selects an appropriate action from the sales strategy action space, such as adjusting airfares or flight schedules. After selecting this action, the agent uses the ticket sales trend analysis module to predict future sales volume for each class, thereby updating the economic state space. Then, based on the sales strategy value function, the agent calculates the immediate reward for taking this action—the profit value calculated based on the current economic state and the sales strategy action taken. This reward value guides the agent's behavioral choices, thereby optimizing the sales strategy. Throughout this process, the agent utilizes a reinforcement learning algorithm to gradually adjust its strategy by combining trial and error with the currently known optimal strategy, aiming to maximize the cumulative ticket profit index.

[0086] By setting up an economic state space, a sales strategy action space, and a sales strategy value function, the module achieves dynamic optimization and fine-tuning of ticket sales strategies. By analyzing the economic state space, the module can monitor sales trends for different cabin classes in real time and adjust sales strategies accordingly to market changes, thereby improving flight profitability and resource utilization efficiency. Furthermore, by incorporating factors such as fuel prices and flight distance, sales strategies are more aligned with actual operating conditions, effectively reducing costs and increasing revenue.

[0087] Furthermore, the adjusted ticket profit index is calculated, and the sales strategy value function is updated through the reinforcement learning algorithm to determine the optimal sales strategy; Figure 2The specific implementation process of step S5 includes:

[0088] After adjusting the sales strategy, the ticket profit index is calculated. The sales strategy value function is updated through the reinforcement learning algorithm to obtain the optimal sales strategy. The specific formula for updating the sales strategy value function is:

[0089]

[0090] R u =es u ·ep u +bs u bp u +fs u ·fp u -(es yc ·ep ys +bs yc bp ys +fs yc ·fp ys )+l u Δop;

[0091] Among them, Q(S u ,A u ) represents the current economic state S u Next, select Sales Strategy Action A u The value of when, Q() is the sales strategy value function; γ represents the learning rate and 0<γ<1; R u Represents the current economic state S u Next, select Sales Strategy Action A u The instant reward is the air ticket profit index; δ represents the discount factor and 0<δ<1; Indicates the next economic state S u+1 The maximum value of the next value, Indicates the next economic state S u+1 Sales strategy action when the next value is the largest; u ,bs u and fs u Respectively represent the selection of sales strategy action A u The future sales volume of economy class, business class and first class at the time of ep u , bp u and fp u Respectively represent the selection of sales strategy action A u Economy class airfare, business class airfare and first class airfare at that time; yc ,bs yc and fs ycThey represent the future sales volume of economy class, business class and first class when no sales strategy adjustment is made; ep ys , bp ys and fp ys Economy class airfare, business class airfare and first class airfare before sales strategy adjustment; u Represents the current economic state S u The flight distance under the current economic status S u The difference between the oil price under the current economic state and the oil price under the previous economic state.

[0092] Optimizing sales strategies through reinforcement learning algorithms maximizes the ticket profit index, thereby improving profitability. Dynamic adjustments to sales strategies based on multiple factors, including the current economic situation, market demand, and fuel prices, ensure maximum ticket sales profits in varying economic environments. By incorporating information such as instant rewards, discount factors, and future sales volumes, precise adjustments can be made to sales strategies for economy, business, and first class, achieving accurate pricing and space allocation, thereby improving overall operational efficiency.

[0093] Example 2

[0094] In an embodiment of the present application, a deep learning analysis system for ticket sales trends is applied to the flight control system of Airline A to analyze historical ticket sales data in order to effectively control flight schedules and ticket fares.

[0095] Specifically, the ticket-related data collection module collects ticket sales data, operation data, and passenger data of Airline A over a historical period, and performs data labeling and data cleaning to obtain historical ticket sales data, historical operation data, and historical passenger data.

[0096] Furthermore, the air ticket-related data processing module obtains the air ticket attribute correlation based on the air ticket attributes of the historical air ticket sales data and the historical sales data of different cabin classes, and determines the key air ticket attributes of different cabin classes by combining the constructed sales volume-air ticket attribute regression model; determines the key operational attributes of different cabin classes by constructing a sales volume-operation attribute regression model and operational attribute screening; calculates the passenger class probability index of different passengers, divides the historical data of passengers of different cabin classes, and combines the data of corresponding key air ticket attributes and key operational attributes to obtain historical sales data sets of different cabin classes.

[0097] Furthermore, the ticket sales trend analysis module constructs a sales volume forecasting model, performs model training and optimization on historical sales data sets of different cabin classes, obtains a trained sales volume forecasting model, and predicts future sales volumes of different cabin classes.

[0098] Furthermore, the air ticket sales strategy control module sets the economic state space, sales strategy action space and sales strategy value function, and adjusts the sales strategy based on the economic state space; calculates the adjusted air ticket profit index, and updates the sales strategy value function through the reinforcement learning algorithm to determine the optimal sales strategy;

[0099] Specifically, when adjusting the sales strategy in the sales strategy action space, the future sales volume of different cabin classes is re-forecasted through the ticket sales trend analysis module to update the economic state space;

[0100] Among them, the economic status space includes future sales volume of economy class, future sales volume of business class, future sales volume of first class, oil prices and flight distance; the sales strategy action space includes air ticket fare adjustments, flight schedule adjustments and aircraft type adjustments.

[0101] Furthermore, the specific implementation process of calculating the adjusted ticket profit index, updating the sales strategy value function, and determining the optimal sales strategy includes:

[0102] After adjusting the sales strategy, the ticket profit index is calculated. The sales strategy value function is updated through the reinforcement learning algorithm to obtain the optimal sales strategy. The specific formula for updating the sales strategy value function is:

[0103]

[0104] R u =es u ·ep u +bs u bp u +fs u ·fp u -(es yc ·ep ys +bs yc bp ys +fs yc ·fp ys )+l u Δop;

[0105] Among them, Q(S u ,A u ) represents the current economic state S u Next, select Sales Strategy Action A u The value of when, Q() is the sales strategy value function; γ represents the learning rate and 0<γ<1; R u Represents the current economic state S u Next, select Sales Strategy Action A u The instant reward is the air ticket profit index; δ represents the discount factor and 0<δ<1; Indicates the next economic state Su+1 the maximum value of the lower value, represents the next economic state S u+1 the sales strategy action when the lower value is maximum; es u , bs u and fs u respectively represent the future sales of economy class, the future sales of business class and the future sales of first class when the sales strategy action A u is selected; ep u , bp u and fp u respectively represent the ticket price of economy class, the ticket price of business class and the ticket price of first class when the sales strategy action A u is selected; es yc , bs yc and fs yc respectively represent the future sales of economy class, the future sales of business class and the future sales of first class when no sales strategy adjustment is made; ep ys , bp ys and fp ys respectively represent the ticket price of economy class, the ticket price of business class and the ticket price of first class when no sales strategy adjustment is made; l u represents the flight distance under the current economic state S u ; Δop represents the difference between the oil price under the current economic state S u and the oil price under the previous economic state.

[0106] The embodiments of the present application can accurately identify key ticket attributes and operation attributes of different passenger cabin levels through in-depth analysis of historical ticket sales data, operation data and passenger data, and effectively predict sales trends using deep learning and regression models. Among them, combined with deep learning algorithm and multi-dimensional feature extraction, the future sales of different passenger cabin levels can be accurately predicted, and key ticket attributes, operation attributes and passenger behavior patterns are extracted according to historical data, to provide accurate sales prediction for airlines. At the same time, the reinforcement learning algorithm is used to dynamically adjust the sales strategy, maximize the ticket profit index, and optimize resource allocation and improve operation efficiency by adjusting the ticket price, flight frequency and aircraft type.

[0107] Although embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and alterations can be made thereto without departing from the principles and spirit of the present application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A deep learning analysis system for air ticket sales trends, characterized by: include: The ticket-related data collection module collects the airline's historical ticket sales data, operation data, and passenger data, and performs data labeling and data cleaning to obtain historical ticket sales data, historical operation data, and historical passenger data; The ticket-related data processing module obtains ticket attribute correlations based on the ticket attributes of the historical ticket sales data and the historical sales data of different cabin classes, and determines the key ticket attributes of different cabin classes by combining the sales volume-ticket attribute regression model; determines the key operational attributes of different cabin classes by building a sales volume-operation attribute regression model and performing operational attribute screening; calculates the passenger class probability index for different passengers, divides the historical passenger data of different cabin classes, and combines the data corresponding to the key ticket attributes and key operational attributes to obtain historical sales data sets for different cabin classes. The specific implementation process of the ticket-related data processing module includes: Dividing the sales volume in the historical ticket sales data according to cabin class to obtain historical sales volume of economy class, historical sales volume of business class, and historical sales volume of first class; For the historical ticket sales data, the ticket attribute correlation between different ticket attributes is calculated using the Pearson correlation coefficient; the sales volume-ticket attribute regression model is constructed, and the degree of influence of different ticket attributes on the historical sales volume of different cabin classes is determined by fitting the historical data, thereby obtaining the influence of economy class ticket attributes, business class ticket attributes, and first class ticket attributes; based on the ticket attribute correlation, the ticket attributes corresponding to different cabin classes are screened to determine the key ticket attributes for economy class, business class, and first class. For the historical operating data, construct the sales volume-operation attribute regression model, and determine the key operating attributes of economy class, business class, and first class by fitting the historical data and screening the operation attributes; For the historical passenger data, the passenger class probability index is calculated based on cabin class, ticket purchase frequency, and ticket purchase purpose. The specific calculation formula is: Among them, PC i represents the passenger class probability index of passenger i; Indicates cabin class sequence C i The value of cabin class c when it appears the most times, f() is the frequency function; exp() represents the exponential function with a natural constant as the base; β represents the ticket purchase frequency sequence F i Average frequency of ticket purchases The exponential weight of the ticket; p represents the ticket purchase purpose sequence P i The specific code value of the ticket purchase purpose; α1 represents the cabin class sequence C of passenger i i The impact factor of α2 represents the ticket purchase frequency sequence F of passenger i. i The impact factor of α3 represents the ticket purchase purpose sequence P of passenger i. i The impact factor is 0<α2<α3<α1<1 and α1+α2+α3=1; dividing the historical passenger data into economy class passenger historical data, business class passenger historical data, and first class passenger historical data according to the passenger class probability index; The historical passenger data, key ticket attributes, and key operational attribute data corresponding to different cabin classes are integrated to obtain the economy class historical sales dataset, the business class historical sales dataset, and the first class historical sales dataset; The ticket sales trend analysis module builds a sales volume forecasting model, trains and optimizes the model on historical sales data sets of different cabin classes, obtains the trained sales volume forecasting model, and predicts future sales volumes of different cabin classes; The air ticket sales strategy control module sets an economic state space, a sales strategy action space, and a sales strategy value function, and adjusts the sales strategy based on the economic state space; calculates the adjusted air ticket profit index, updates the sales strategy value function through a reinforcement learning algorithm, and determines the optimal sales strategy; the specific implementation process of calculating the adjusted air ticket profit index, updating the sales strategy value function, and determining the optimal sales strategy includes: After adjusting the sales strategy, the ticket profit index is calculated; the sales strategy value function is updated through the reinforcement learning algorithm to obtain the optimal sales strategy; the specific formula for updating the sales strategy value function is: R u =es u ·ep u +bs u ·bp u +fs u ·fp u -(es yc ·ep ys +bs yc ·bp ys +fs yc ·fp ys )+l u ·Δop; Among them, Q(S u ,A u ) represents the current economic state S u Next, select Sales Strategy Action A u The value of the sales strategy is, Q() is the sales strategy value function; γ represents the learning rate and 0<γ<1; R u Represents the current economic state S u Next, select Sales Strategy Action A u The instant reward is the ticket profit index; δ represents the discount factor and 0<δ<1; Indicates the next economic state S u+1 The maximum value of the next value, Indicates the next economic state S u+1 Sales strategy action when the next value is the largest; u ,bs u and fs u Respectively represent the selection of sales strategy action A u The future sales volume of economy class, business class and first class at the time of ep u , bp u and fp u Respectively represent the selection of sales strategy action A u Economy class airfare, business class airfare and first class airfare at that time; yc ,bs yc and fs yc They represent the future sales volume of economy class, business class and first class when no sales strategy adjustment is made; ep ys , bp ys and fp ys Economy class airfare, business class airfare and first class airfare before sales strategy adjustment; u Represents the current economic state S u The flight distance under the current economic status S u The difference between the oil price under the current economic state and the oil price under the previous economic state.

2. The deep learning analysis system for air ticket sales trends according to claim 1, characterized in that: The ticket sales data includes the take-off time, landing time, flight duration, date, ticket price and sales volume of each flight; the operational data includes aircraft type, failure rate and cabin service quality; passenger data includes cabin class, ticket purchase frequency and ticket purchase purpose; the data labeling process is to label the corresponding holidays according to the date of each flight, and replace the date attributes in the ticket sales data with holidays.

3. The deep learning analysis system for air ticket sales trends according to claim 1, characterized in that: The sales volume prediction model includes a multi-dimensional feature extraction layer, a feature analysis and fusion layer, and a sales trend prediction layer; The multidimensional feature extraction layer is used to extract ticket sales features, flight operation features, aircraft operation features and passenger features from the historical sales dataset of different cabin classes; The feature analysis and fusion layer is used to analyze the impact of the flight operation characteristics, the aircraft operation characteristics, and the passenger characteristics on the ticket sales characteristics of different cabin classes, and perform feature weighted fusion; The sales trend prediction layer is used to predict the future sales volume of different cabin classes.

4. The deep learning analysis system for air ticket sales trends according to claim 3, characterized in that: The specific implementation process of performing feature analysis and feature weighted fusion through the feature analysis and fusion layer includes: The flight operation characteristics and the passenger characteristics are input into the cross-attention mechanism to capture the potential impact of different flights on passengers' ticket purchasing behavior and obtain the passenger-flight preference weight; the aircraft operation characteristics and the passenger characteristics are input into the cross-attention mechanism to analyze the preferences of different passengers for aircraft operation conditions and obtain the passenger-aircraft preference weight; the passenger characteristics, the aircraft operation characteristics and the flight operation characteristics are weightedly fused in combination with the passenger-flight preference weight and the passenger-aircraft preference weight to obtain the passenger preference characteristics; through the LSTM network, the influence of the passenger preference characteristics on the ticket sales characteristics is explored to obtain the multi-dimensional fusion characteristics of ticket sales.

5. The deep learning analysis system for air ticket sales trends according to claim 1, characterized in that: When adjusting the sales strategy in the sales strategy action space, re-forecasting the future sales volume of different cabin classes through the ticket sales trend analysis module to update the economic state space; The economic status space includes future sales volume of economy class, future sales volume of business class, future sales volume of first class, oil prices and flight distance; the sales strategy action space includes ticket fare adjustment, flight schedule adjustment and aircraft type adjustment.

Images