Intelligent ticketing system

Abstract

The invention discloses an intelligent ticketing system, and relates to the technical field of intelligent ticketing, and the system comprises a data collection module which obtains historical sales, weather forecast and holiday and festival data; the first prediction module constructs a linear fitting function by using the user behavior data and the historical sales data, and predicts the ticket selling quantity of each type of ticket business in the next week in the workday and the dual-holiday; the second prediction module is used for predicting the ticket selling quantity of the next day based on the average ticket selling quantity of the previous four weeks and the weather data, and considering the influence of holidays and festivals; the comprehensive prediction module calculates errors and determines respective weight factors by combining the actual ticket selling quantity of the last week, the first predicted ticket selling quantity and the second predicted ticket selling quantity, and the comprehensive predicted ticket selling quantity of all types of tickets in the next day is obtained; a ticket issuing channel adjusting module dynamically adjusts the number of ticket issuing channels according to the predicted ticket selling quantity and a single-channel ticket issuing threshold value, and optimizes distribution of new ticket issuing requests so as to prevent network congestion and improve ticket issuing efficiency; and the ticket issuing optimization operation of the intelligent ticket business is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent ticketing, and in particular to an intelligent ticketing system. Background Art

[0002] Smart ticketing system is an innovative ticketing management system that uses advanced technology and information technology, based on wireless communication, data transmission, mobile payment and other technologies, to transform traditional physical forms such as cards and tickets into electronic or virtual forms;

[0003] As an important tool for modern performance and event management, the accuracy and efficiency of the smart ticketing system are crucial to improving user experience, optimizing resource allocation, and increasing revenue. With the rapid development of big data and artificial intelligence technologies, the smart ticketing system can not only process a large amount of ticketing data, but also predict future ticket sales trends by analyzing these data, thereby providing support for decision-making.

[0004] However, existing intelligent ticketing systems usually predict ticket sales based on simple historical data analysis. This method often ignores the impact of multiple factors such as user behavior, weather changes, and holidays, resulting in a large deviation between the prediction results and the actual ticket sales situation. In addition, existing ticketing systems usually only consider adding ticket processing channels when network congestion occurs. They are unable to reasonably allocate ticket processing channels based on the prediction of the next day's ticket sales, and are unable to prevent network congestion in advance. This leads to processing delays when congestion occurs, increases user waiting time, and reduces system efficiency. Summary of the invention

[0005] 1. Technical issues to be solved

[0006] In response to the technical problems in the background technology, the present invention proposes an intelligent ticketing system, comprising multiple modules, specifically: a data acquisition module obtains historical sales, weather forecast and holiday data; a first prediction module uses user behavior data and historical sales data to construct a linear fitting function to predict the ticket sales volume of each type of ticket on weekdays and weekends in the next week; the second prediction module predicts the ticket sales volume of the next day based on the average ticket sales volume and weather data of the previous four weeks, and takes into account the impact of holidays; the comprehensive prediction module combines the actual ticket sales volume of the previous week, the first prediction and the second prediction ticket sales volume, calculates the error and determines the respective weight factors, and obtains the comprehensive predicted ticket sales volume of each type of ticket on the next day; the ticket issuance channel adjustment module dynamically adjusts the number of ticket issuance channels according to the predicted ticket sales volume and the single-channel ticket issuance threshold, and optimizes the allocation of new ticket issuance requests; thereby solving the problems mentioned in the background technology.

[0007] (II) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] An intelligent ticketing system, comprising:

[0010] The data collection module obtains the sales data of various types of tickets from the ticket database of the intelligent ticketing system at the end of each day, including the sales record data of various types of tickets on the same day of the next week in the previous four weeks, and obtains the weather forecast data and holiday data for the next day;

[0011] The first prediction module calculates each user's score reliability index Sri according to the actual number of times each user purchases tickets in the intelligent ticketing system. k , and combined with the rating data of users on various types of tickets in the previous week, calculate the comprehensive rating Os of various types of tickets in the previous week i , construct a linear fitting function of the comprehensive score of each type of ticket and the actual ticket sales in the first four weeks, and calculate the first predicted ticket sales Fts of each type of ticket in the next week i For the working days and weekends of the next week, the average ratio of ticket sales of various types of ticket working days to weekends in the previous four weeks is used. i Calculate the first predicted ticket sales volume for each type of ticket for each working day and each weekend of the next week; for the holidays of the next week, construct a linear fitting function between the comprehensive scores of each type of ticket and the actual ticket sales volume for the corresponding holidays in the previous four years, and obtain the first predicted ticket sales volume for the holiday based on the comprehensive ticket score data for the corresponding holiday in the previous year;

[0012] The second prediction module predicts the ticket sales of each type of ticket on the next day based on the average ticket sales data of each type of ticket in the same week as the next day in the previous four weeks and the average ticket sales data under various weather conditions, and obtains the second predicted ticket sales Ets of each type of ticket on the next day i If the next day is a holiday, the average ticket sales of each type of ticket during the holiday in the past three years is combined with the weather forecast data of the next day to predict the ticket sales of each type of ticket on the next day;

[0013] Comprehensive forecasting module, based on the actual ticket sales data of various types of tickets per day in the previous week Ats io , and the first predicted ticket sales volume Yts for each type of ticket io and the second predicted ticket sales Ets io , respectively calculate the first error Fe of the first predicted ticket sales volume and the second predicted ticket sales volume relative to the actual ticket sales volume data io and the second error Se io , according to the first error Fe io and the second error Se ioGet the weight factor β of the first predicted ticket sales volume and the second predicted ticket sales volume 1 With β 2 ; The first predicted ticket sales volume Yts for each type of ticket for the next day io and the second predicted ticket sales Ets io Combined with the corresponding weight factor, the comprehensive forecast ticket sales volume Cts of each type of ticket for the next day is obtained i ;

[0014] The ticket channel adjustment module simulates the network status of each type of ticket when different numbers of tickets are issued in a single channel, and obtains the single channel ticket threshold St for each type of ticket. i ; Based on the comprehensive forecast of ticket sales volume Cts for each type of ticket i The single-channel ticket volume threshold St i The relationship between the number of ticketing channels for each type of ticket is adjusted; when a new ticketing request appears in the ticketing waiting queue, it is determined whether the number of ticketing channels needs to be adjusted.

[0015] Specifically, at the end of each day, the ticket sales data of each type of ticket in the same week as the next day in the previous four weeks is obtained from the ticket database, recorded as Pw ij , where j represents the previous jth week; obtain the weather forecast data for the next day, and use We to represent the weather forecast data for the next day. When We=1, it means it is a rainy or snowy day, and when We=2, it means it is a non-rainy or snowy day.

[0016] Specifically, count the actual number of ticket purchases by all users in the smart ticketing system Atp k , where k represents the kth user, and the actual number of ticket purchases represents the number of times a user purchases tickets normally and does not refund them; according to the actual number of ticket purchases Atp by all users in the smart ticketing system k , calculate the rating reliability index Sri for each user k , the expression is:

[0017]

[0018] Among them, Mi k is the member index, and Mi k ∈{1, N}, and N>1, when the user is a non-member user, Mi k =1; when the user is a member user, Mi k =N; the value of Max{x} is the maximum value of x;

[0019] Get the ticket sales data of various types of tickets in the previous week, and get the names of the users who purchased the tickets of various types of tickets. According to the names of the users who purchased the tickets, get the user's rating data for each type of ticket in the corresponding user table. ikand each user's rating reliability index Sri k Combined to get the comprehensive score Os of various types of tickets in the previous week i , the expression is:

[0020]

[0021] Among them, m i Represents the total number of tickets actually purchased for the i-th type of ticket.

[0022] Further, the comprehensive score data of each type of ticket in the second week, the third week and the fourth week, as well as the actual ticket sales data of each type of ticket in the first week, the second week and the third week are obtained from the weekly record table in the ticket database; based on the linear regression equation Y ij =b+a*X i(j+1) The comprehensive score data of each type of ticket in the second week, the third week, and the fourth week, as well as the actual ticket sales data of each type of ticket in the first week, the second week, and the third week are fitted, where Y ij represents the actual sales volume of the ith type of ticket in the previous j week, X i(j+1) Represents the comprehensive score of the i-th type of ticket in the previous j+1 week; the slope a of the fitting line is calculated by the least squares method o and the intercept b 0 , the fitting equation is Y ij =b 0 +a 0 *X i(j+1) .

[0023] Furthermore, the comprehensive score data Os of each type of ticket in the previous week, i.e. the previous week, is i Substituting into the fitting equation, we can get the first predicted ticket sales volume Fts for each type of ticket in the next week. i ; Filter out the ticket sales of each type of ticket on weekdays and weekends in the first week, the second week, the third week, and the fourth week, and calculate the ticket sales of each type of ticket on weekdays Wts in each week ij And the weekend ticket sales volume Sts ij , through the formula Calculate the average ratio of ticket sales on weekdays to weekends for each type of ticketing i , where n represents the number of weeks without holidays in the first week, the second week, the third week, and the fourth week, and n∈{1,…,4}; the first predicted ticket sales volume Fts of each type of ticket in the next week i The average ratio of ticket sales on weekdays to weekends for all types of ticket sales i Combined, we get the first predicted total ticket sales volume Wts for each type of ticket on the next weekday i, and the first forecast total ticket sales volume of various types of tickets on weekends Sts i ; Then get the first predicted ticket sales volume Yts for each day of the next week i , the first predicted ticket sales volume for each type of ticket for the next weekday For the weekends in the next week, the first forecast ticket sales of each type of ticket

[0024] For holidays within the next week, obtain the name of the holiday and obtain the ticket sales data of each type of ticket within the previous year, the previous year, the previous year, the previous year, and the previous year from the ticket database, and obtain the names of the users who purchased each type of ticket in each year, obtain the user's rating data for each type of ticket in the corresponding user table according to the name of the ticket purchasing user, and calculate the comprehensive ratings of each type of ticket within the previous year, the previous year, the previous year, and the previous year through the comprehensive rating calculation formula; combine the comprehensive ratings of each type of ticket within the previous year, the previous year, and the previous year with the actual ticket sales data of each type of ticket within the previous year, the previous year, and the previous year, respectively, to construct a linear fitting function for the corresponding holiday, and bring the comprehensive ratings of each type of ticket within the previous year into the linear fitting function to obtain the first predicted ticket sales volume Yts for the holiday in the next week. i .

[0025] Specifically, at the end of each day, the ticket sales data of each type of ticket in the previous four weeks are obtained from the ticket database, and the ticket sales data of each type of ticket in the same week in the previous four weeks are combined to obtain the average ticket sales data of each type of ticket on each day XAts i(week) , where week=1 means Monday, and so on; obtain the ticket sales data under various weather conditions from the ticket database, combine the ticket sales data for each type of weather, and obtain the average ticket sales data for each type of ticket under each type of weather WAts i(we) ;

[0026] Get the next day's week information and weather forecast data; if the next day is not a holiday, combine the average ticket sales data of each type of ticket for the next day's corresponding week and the average ticket sales data of each type of ticket for the corresponding weather to get the second predicted ticket sales volume Ets for each type of ticket for the next day i , the expression is:

[0027] Ets i =α 1 *XAts i(week) +α 2 *WAts i(we)

[0028] Among them, α 1, α 2 are the weight coefficients of week information and weather information respectively;

[0029] Furthermore, if the next day is a holiday, obtain the name of the holiday and obtain the ticket sales data of each type of ticket during the holiday in the past three years from the ticket database. Combine the obtained ticket sales data of each type of ticket during the three holidays to obtain the average ticket sales of each type of ticket during the holiday HAts i(ho) , where ho represents different holidays; the average ticket sales of each type of ticket on the next day corresponding to the holiday is combined with the average ticket sales of each type of ticket on the corresponding weather to obtain the second predicted ticket sales of each type of ticket on the next day Ets i , the expression is: Ets i =α 1 *HAts i(ho) +α 2 *WAts i(we) .

[0030] Specifically, obtain the actual ticket sales data of each type of ticket every day in the previous week from the ticket database. io , and the first predicted ticket sales volume Yts for each type of ticket every day in the previous week io and the second predicted ticket sales Ets io , where o represents the oth day in the previous week; the actual ticket sales data Ats io and the first predicted ticket sales volume Yts io and the second predicted ticket sales Ets io After combining, we get the first error Fe between the first predicted ticket sales volume and the second predicted ticket sales volume relative to the actual ticket sales volume data. io and the second error Se io , the expression is Fe io =|Ats io -Yts io |,Se io =|Ats io -Ets io |; The first daily error Fe in the previous week io and the second error Se io Find the average and get the first average error and the second mean error

[0031] The first average error The second mean error Combined, calculate the weight factor β of the first predicted ticket sales volume and the second predicted ticket sales volume respectively 1 With β 2 , the expression is: Combine the first predicted ticket sales volume and the second predicted ticket sales volume of each type of ticket for the next day with the corresponding weight factor to obtain the comprehensive predicted ticket sales volume Cts of each type of ticket for the next day i , the expression is: Cts i =β 1 *Yts i +β 2 *Ets i .

[0032] Specifically, a simplified model of the intelligent ticketing system is built, and the ticket issuance requests corresponding to each type of ticket are simulated in the simplified model to perform ticket issuance processing in a single-channel manner; ticket issuance requests are sent to the simplified model of the intelligent ticketing system one by one, and the network status data of the intelligent ticketing system is monitored in real time. The real-time network status data is analyzed by the network analysis tool in the system. When the network status is first displayed as congested, the total number of ticket issuance requests at this time is recorded, and the total number of ticket issuance requests obtained by simulating each type of ticket when the network status is first displayed as congested is used as the single-channel ticket issuance threshold St for the corresponding type of ticket. i .

[0033] Furthermore, the comprehensive forecast ticket sales volume Cts of each type of ticket i The single-channel ticket volume threshold St i Combined, if Cts i <St i , two channels are allocated for this type of ticket to process the ticket; if Cts i ≥St i , then set it as the ticket allocation of this type +2 channels for ticket processing, including Indicates that x is rounded down;

[0034] Whenever a new ticket waiting request appears in the ticket waiting queue of each channel, the total number of ticket waiting requests Npr in the ticket waiting queue of the channel is counted. i , if Npr i ≥0.9*St i , you need to use a new channel for ticketing operations. Specifically, take out the ticketing waiting requests sorted as even numbers in the ticketing waiting queue of the current channel, put them into the ticketing waiting queue of a new channel, and perform subsequent ticketing processing operations.

[0035] (III) Beneficial effects

[0036] The present invention provides an intelligent ticketing system, which has the following beneficial effects:

[0037] 1. The data collection module automatically extracts sales data of various types of tickets from the ticket database at the end of each day, and combines it with weather forecasts and holiday information. This module provides a solid foundation for subsequent predictions and analysis;

[0038] 2. The first prediction module builds a more refined prediction model by deeply mining user behavior data and combining the user's ticket purchase frequency, rating, and membership status. It not only takes into account the user's historical ticket purchase behavior, but also measures the influence of user ratings through the rating reliability index, thereby more accurately predicting the ticket sales of various types of tickets in the next week.

[0039] 3. The second prediction module provides another dimension of prediction for the ticket sales of various types of tickets for the next day by comprehensively considering the week information, weather information, and the special impact of holidays. The entropy weight method is used to objectively assign weights to different influencing factors, making the prediction results more accurate and reliable.

[0040] 4. The comprehensive prediction module integrates the prediction results of the first prediction module and the second prediction module, and combines the actual ticket sales data of the previous week to calculate the errors of the two prediction modules and assign weight factors accordingly; this not only improves the accuracy of the prediction, but also enhances the stability of the prediction results;

[0041] 5. The ticketing channel adjustment module obtains the single-channel ticketing threshold for each type of ticket by simulating the network status when processing different numbers of tickets for each type of ticket in a single channel; it can adjust the number of ticketing channels in advance according to the predicted ticket sales volume, effectively preventing network congestion caused by too many ticketing requests; by real-time monitoring of the network status and dynamically adjusting the ticketing channels, the module ensures the efficient operation of the intelligent ticketing system, improves the user experience, and reduces potential losses caused by network congestion. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a schematic diagram of the structure of the intelligent ticketing system provided by the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0044] refer to Figure 1 The present invention provides an intelligent ticketing system, comprising:

[0045] The data collection module obtains the sales data of various types of tickets from the ticket database of the intelligent ticketing system at the end of each day, including the sales record data of various types of tickets on the same day of the next week in the previous four weeks, and obtains the weather forecast data and holiday data for the next day;

[0046] The ticket database is the core component of the intelligent ticketing system. It stores various information and data related to ticket sales. The intelligent ticketing system sells multiple types of tickets. The ticket database contains ticket tables for each type of ticket. Each ticket table stores daily historical sales record data for the corresponding type of ticket and data such as the weather conditions at the start of the performance corresponding to each type of ticket. After each week, the analysis data for each week is stored in the weekly record table. The analysis data includes the total ticket purchase data for each week, the comprehensive score of each type of ticket, and other data. The ticket database also contains a forecast ticket sales table for each type of ticket, which is used to store all forecast data for the next day's ticket sales.

[0047] The user database is a key component of the smart ticketing system. It stores various information and data related to user behavior. The user database includes user tables of all users who have registered accounts in the smart ticketing system. When a user registers an account in the smart ticketing system, a new user table is created in the user database. Each user table stores relevant data of each user, including whether he is a member, the ticket purchase data of each history, as well as the ticket refund record and the rating record of the corresponding performance of the ticket. Among them, using Sc i represents the user's rating of the performance corresponding to the i-th type of ticket office, and Sc i ∈{0, 1, …, 10};

[0048] The ticket database and the user database are connected through the two keywords of user name and ticket type;

[0049] At the end of each day, the ticket sales data of each type of ticket in the same week as the next day in the previous four weeks are obtained from the ticket database, recorded as Pw ij , where j represents the jth week before. Specifically, if tomorrow is Tuesday, the ticket sales data for each Tuesday in the previous four weeks is obtained from the ticket database to obtain the weather forecast data for the next day. We is used to represent the weather forecast data for the next day. When We=1, it represents a rainy or snowy day. When We=2, it represents a non-rainy or snowy day. The weather affects people's travel willingness, the operating conditions of transportation vehicles, safety factors and other aspects. The forecast weather specifically represents the forecast weather conditions at the start of the performance corresponding to each type of ticket.

[0050] The data collection module automatically extracts sales data of various types of tickets from the ticket database at the end of each day, and combines it with weather forecasts and holiday information. This module provides a solid foundation for subsequent predictions and analysis.

[0051] The first prediction module calculates each user's score reliability index Sri according to the actual number of times each user purchases tickets in the intelligent ticketing system. k , and combined with the rating data of users on various types of tickets in the previous week, calculate the comprehensive rating Os of various types of tickets in the previous week i , construct a linear fitting function of the comprehensive score of each type of ticket and the actual ticket sales in the first four weeks, and calculate the first predicted ticket sales Fts of each type of ticket in the next week i For the working days and weekends of the next week, the average ratio of ticket sales of various types of ticket working days to weekends in the previous four weeks is used. i Calculate the first predicted ticket sales volume for each type of ticket for each working day and each weekend of the next week; for the holidays of the next week, construct a linear fitting function between the comprehensive scores of each type of ticket and the actual ticket sales volume for the corresponding holidays in the previous four years, and obtain the first predicted ticket sales volume for the holiday based on the comprehensive ticket score data for the corresponding holiday in the previous year;

[0052] In the first prediction module, the following steps are performed:

[0053] S1. Count the actual number of ticket purchases by all users in the smart ticketing system Atp k , where k represents the kth user, and the actual number of ticket purchases represents the number of times a user purchases tickets normally and does not refund them; according to the actual number of ticket purchases Atp by all users in the smart ticketing system k , calculate the rating reliability index Sri for each user k , the expression is:

[0054]

[0055] Among them, Mi k is the member index, and Mi k ∈{1, N}, and N>1, when the user is a non-member user, Mi k =1; when the user is a member user, Mi k =N, where the value of N is set by the system administrator. The ratings of member users have a higher influence than those of non-member users. The value of Max{x} is the maximum value of x.

[0056] S2. Obtain the ticket sales data of various types of tickets in the previous week, and obtain the names of the users who purchased the tickets of various types of tickets. According to the names of the users who purchased the tickets, obtain the user's rating data Sc for each type of ticket in the corresponding user table. ikand each user's rating reliability index Sri k Combined to get the comprehensive score Os of various types of tickets in the previous week i , the expression is:

[0057]

[0058] Among them, m i represents the total number of actual tickets purchased for the ith type of ticket; combining the user's rating data for each type of ticket with each user's rating reliability index is used to more accurately predict the ticket sales volume for the next week from the user's psychological perspective;

[0059] S3. Obtain the comprehensive scoring data of each type of ticket in the second week, the third week and the fourth week, and the actual ticket sales data of each type of ticket in the first week, the second week and the third week from the weekly record table in the ticket database; the comprehensive scoring data of each type of ticket in the second week is used to predict the ticket sales of each type of ticket in the first week, the comprehensive scoring data of each type of ticket in the third week is used to predict the ticket sales of each type of ticket in the second week, and the comprehensive scoring data of each type of ticket in the fourth week is used to predict the ticket sales of each type of ticket in the third week;

[0060] Based on the linear regression equation Y ij =b+a*X i(j+1) The comprehensive score data of each type of ticket in the second week, the third week, and the fourth week, as well as the actual ticket sales data of each type of ticket in the first week, the second week, and the third week are fitted, where Y ij represents the actual sales volume of the ith type of ticket in the previous j week, X i(j+1) Represents the comprehensive score of the i-th type of ticket in the previous j+1 week; the slope a of the fitting line is calculated by the least squares method 0 and the intercept b 0 , the fitting equation is Y ij =b 0 +a 0 *X i(j+1) ;

[0061] S4. The comprehensive score data of various types of tickets in the previous week, i.e., the previous week, Os i Substituting into the fitting equation, we can get the first predicted ticket sales volume Fts for each type of ticket in the next week. i ; Filter out the ticket sales of various types of tickets on weekdays (Monday to Friday) and weekends (Saturday and Sunday) in the first, second, third and fourth weeks before the previous week, and calculate the weekday ticket sales of each type of ticket Wts for each week ij And the weekend ticket sales volume Sts ij , through the formula Calculate the average ratio of ticket sales on weekdays to weekends for each type of ticketing i , where n represents the number of weeks without holidays (including Chinese statutory holidays and Western statutory holidays) in the first week, the second week, the third week, and the fourth week, and n∈{1,…,4}; the first predicted ticket sales volume Fts of each type of ticket in the next week i The average ratio of ticket sales on weekdays to weekends for all types of ticket sales i Combined, we get the first predicted total ticket sales volume Wts for each type of ticket on the next weekday i , and the first forecast total ticket sales volume of various types of tickets on weekends Sts i ; Then get the first predicted ticket sales volume Yts for each day of the next week i , the first predicted ticket sales volume for each type of ticket for the next weekday For the weekends in the next week, the first forecast ticket sales of each type of ticket The working days and weekends mentioned do not include holidays;

[0062] For holidays within the next week, obtain the name of the holiday and obtain the ticket sales data of each type of ticket within the previous year, the previous year, the previous year, the previous year, and the previous year from the ticket database, and obtain the names of the users who purchased each type of ticket in each year, and obtain the user's rating data for each type of ticket in the corresponding user table according to the name of the ticket purchasing user, and calculate the comprehensive scores of each type of ticket in the holiday in the previous year, the previous year, the previous year, and the previous year through the comprehensive score calculation formula in S2; combine the comprehensive scores of each type of ticket in the holiday in the previous year, the previous year, and the previous year with the actual ticket sales data of each type of ticket in the holiday in the previous year, the previous year, and the previous year, respectively, to construct a linear fitting function for the corresponding holiday, and bring the comprehensive scores of each type of ticket in the holiday in the previous year into the linear fitting function to obtain the first predicted ticket sales volume Yts for the holiday in the next week. i .

[0063] The first prediction module builds a more refined prediction model by deeply mining user behavior data and combining the user's ticket purchase frequency, ratings, and membership status. It not only takes into account the user's historical ticket purchase behavior, but also measures the influence of user ratings through the rating reliability index, thereby more accurately predicting the sales volume of various types of tickets in the next week.

[0064] The second prediction module predicts the ticket sales of each type of ticket on the next day based on the average ticket sales data of each type of ticket in the same week as the next day in the previous four weeks and the average ticket sales data under various weather conditions, and obtains the second predicted ticket sales Ets of each type of ticket on the next dayi If the next day is a holiday, the average ticket sales of each type of ticket during the holiday in the past three years is combined with the weather forecast data of the next day to predict the ticket sales of each type of ticket on the next day;

[0065] In the second prediction module, the following steps are performed:

[0066] T1. At the end of each day, obtain the ticket sales data of each type of ticket for the previous four weeks from the ticket database, and combine the ticket sales data of each type of ticket for the same week in the previous four weeks to obtain the average ticket sales data of each type of ticket for each day XAts i(week) , where week=1 means Monday, and so on; for example: obtain the ticket sales data of each type of ticket on Wednesday in the first week, the second week, the third week, and the fourth week respectively, and combine the ticket sales data of each type of ticket on all Wednesdays in the previous four weeks to obtain the average ticket sales data WAts on Wednesday 3 ; Get the ticket sales data for each type of weather from the ticket database, combine the ticket sales data for each type of weather, and get the average ticket sales data for each type of ticket under each type of weather. i(we) ; These data need to be updated at the end of each day;

[0067] T2. At the end of each day, obtain the next day's week information, that is, determine the next day's week, and obtain the next day's weather forecast data; determine whether the next day is a holiday. If the next day is not a holiday, combine the average ticket sales data of each type of ticket corresponding to the week of the next day and the average ticket sales data of each type of ticket corresponding to the weather to obtain the second predicted ticket sales volume Ets of each type of ticket for the next day. i , the expression is:

[0068] Ets i =α 1 *XAts i(week) +α 2 *WAts i(we)

[0069] Among them, α 1 , α 2 are the weight coefficients of the day of the week information and the weather information, respectively. The entropy weight method is used to calculate the weight coefficients of the day of the week information and the weather information. The entropy weight method is an objective weighting method based on information theory. It determines the weight by calculating the order degree (i.e., the degree of dispersion) of the information contained in each indicator. In multi-attribute decision-making problems, especially when multiple indicators need to be considered comprehensively (such as the day of the week information and weather information in the present invention), the entropy weight method can provide a more objective weight distribution.

[0070] T3. If the next day is a holiday, obtain the name of the holiday and obtain the ticket sales data of each type of ticket during the holiday in the past three years from the ticket database. Combine the three obtained ticket sales data of each type of ticket during the holiday to obtain the average ticket sales of each type of ticket during the holiday. i(ho) , where ho represents different holidays and is represented by different numbers; the average ticket sales of each type of ticket corresponding to the holiday on the next day is combined with the average ticket sales of each type of ticket corresponding to the weather to obtain the second predicted ticket sales of each type of ticket on the next day Ets i , the expression is: Ets i =α 1 *HAts i(ho) +α 2 *WAts i(we) .

[0071] The second prediction module provides another dimension of prediction for the ticket sales volume of various types of tickets for the next day by comprehensively considering the week information, weather information, and the special impact of holidays; the entropy weight method is used to objectively assign weights to different influencing factors, making the prediction results more accurate and reliable.

[0072] Comprehensive forecasting module, based on the actual ticket sales data of various types of tickets per day in the previous week Ats io , and the first predicted ticket sales volume Yts for each type of ticket io and the second predicted ticket sales Ets io , respectively calculate the first error Fe of the first predicted ticket sales volume and the second predicted ticket sales volume relative to the actual ticket sales volume data io and the second error Se io , according to the first error Fe io and the second error Se io Get the weight factor β of the first predicted ticket sales volume and the second predicted ticket sales volume 1 With β 2 ; The first predicted ticket sales volume Yts for each type of ticket for the next day io and the second predicted ticket sales Ets io Combined with the corresponding weight factor, the comprehensive forecast ticket sales volume Cts of each type of ticket for the next day is obtained i ;

[0073] In the comprehensive prediction module, the following steps are performed:

[0074] P1. Get the actual ticket sales data of each type of ticket every day in the previous week from the ticket database. io , and the first predicted ticket sales volume Yts for each type of ticket every day in the previous week io and the second predicted ticket sales Ets io, where o represents the oth day of the previous week; the actual ticket sales data of each type of ticket every day in the previous week Ats io The corresponding first predicted ticket sales volume Yts io and the second predicted ticket sales volume Ets io Combined, the first error Fe between the first predicted ticket sales volume and the second predicted ticket sales volume relative to the actual ticket sales volume data is obtained. io and the second error Se io , the expression is: Fe io =|Ats io -Yts io |,Se io =|Ats io -Ets io |; The first daily error Fe in the previous week io and the second error Se io Find the average and get the first average error and the second mean error

[0075] P2, the first average error The second mean error Combined, calculate the weight factor β of the first predicted ticket sales volume and the second predicted ticket sales volume respectively 1 With β 2 , the expression is:

[0076]

[0077] P3. Combine the first predicted ticket sales volume and the second predicted ticket sales volume of each type of ticket for the next day with the corresponding weight factor to obtain the comprehensive predicted ticket sales volume Cts of each type of ticket for the next day i , the expression is: Cts i =β 1 *Yts i +β 2 *Ets i .

[0078] The comprehensive prediction module integrates the prediction results of the first prediction module and the second prediction module, and combines the actual ticket sales data of the previous week to calculate the error of the two prediction modules and assign weight factors accordingly; this not only improves the accuracy of the prediction, but also enhances the stability of the prediction results.

[0079] The ticket channel adjustment module simulates the network status of each type of ticket when different numbers of tickets are issued in a single channel, and obtains the single channel ticket threshold St for each type of ticket. i ; Based on the comprehensive forecast of ticket sales volume Cts for each type of ticket i The single-channel ticket volume threshold Sti The number of ticketing channels for each type of ticket is adjusted based on the relationship between the ticketing channels and the number of ticketing channels for each type of ticket. When a new ticketing request appears in the ticketing waiting queue, it is determined whether the number of ticketing channels needs to be adjusted.

[0080] Build a simplified model of the intelligent ticketing system, which includes the ticket processing logic and the interactive interface and analysis function of the network; simulate the ticketing requests corresponding to each type of ticket in the simplified model of the intelligent ticketing system to perform ticketing processing in a single-channel manner, and one ticket corresponds to one ticketing request; send ticketing requests to the simplified model of the intelligent ticketing system one by one, and monitor the network status data of the intelligent ticketing system in real time. Analyze the real-time network status data through the network analysis tool in the system. When the network status is first displayed as congested, record the total number of ticketing requests at this time, and use the total number of ticketing requests obtained by simulating each type of ticket when the network status is first displayed as congested as the single-channel ticketing threshold St for the corresponding type of ticket. i ;

[0081] The comprehensive forecast ticket sales volume Cts of each type of ticket i The single-channel ticket volume threshold St i Combined, the comprehensive forecast ticket sales volume is Cts i This means that the total number of predicted ticket requests is Cts i , if Cts i <St i , it means that according to the total number of predicted ticket issuance requests for this type of ticket, regular ticket issuance can be processed through a single channel, and two channels are allocated for this type of ticket for ticket issuance processing, one of which is used to handle special situations, such as the sudden occurrence of actual ticket sales exceeding the single-channel ticket issuance threshold; if Cts i ≥St i , it means that according to the total number of predicted ticketing requests for this type of ticket, it is not possible to process ticketing through a single channel, and multiple channels need to be set up for regular ticketing processing. There are three channels for ticket issuance: Indicates that x is rounded down; adjusting the number of ticketing channels for each type of ticketing in advance based on the comprehensive forecast of ticket sales can prevent network congestion caused by excessive ticketing requests in the future and reduce the time spent on redistribution when congestion occurs;

[0082] Whenever a new ticket waiting request appears in the ticket waiting queue of each channel, the total number of ticket waiting requests Npr in the ticket waiting queue of the channel is counted. i , if Npr i ≥0.9*St i, it means that the total number of ticket waiting requests in the ticket waiting queue of the channel is too large, and a new channel needs to be used for ticketing operations. Specifically, the ticket waiting requests sorted as even numbers in the ticket waiting queue of the current channel are taken out and put into the ticket waiting queue of a new channel, and subsequent ticketing processing operations are performed.

[0083] The ticketing channel adjustment module obtains the single-channel ticketing threshold for each type of ticket by simulating the network status when processing different numbers of tickets for each type of ticket in a single channel; it can adjust the number of ticketing channels in advance according to the predicted ticket sales volume, effectively preventing network congestion caused by too many ticketing requests; by real-time monitoring of the network status and dynamically adjusting the ticketing channels, the module ensures the efficient operation of the intelligent ticketing system, improves the user experience, and reduces potential losses caused by network congestion.

[0084] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer storage medium or transmitted via a computer storage medium.

[0085] Computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. Computer storage media can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media. Available media can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid state drives (SSDs)).

[0086] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An intelligent ticketing system, characterized in that: include: The data collection module obtains the sales data of various types of tickets from the ticket database of the intelligent ticketing system at the end of each day, including the sales record data of various types of tickets on the same day of the next week in the previous four weeks, and obtains the weather forecast data and holiday data for the next day; The first prediction module calculates each user's score reliability index Sri according to the actual number of times each user purchases tickets in the intelligent ticketing system. k , and combined with the rating data of users on various types of tickets in the previous week, calculate the comprehensive rating Os of various types of tickets in the previous week i , construct a linear fitting function of the comprehensive score of each type of ticket and the actual ticket sales in the first four weeks, and calculate the first predicted ticket sales Fts of each type of ticket in the next week i For the working days and weekends of the next week, the average ratio of ticket sales of various types of ticket working days to weekends in the previous four weeks is used. i Calculate the first predicted ticket sales volume for each type of ticket for each working day and each weekend of the next week; for the holidays of the next week, construct a linear fitting function between the comprehensive scores of each type of ticket and the actual ticket sales volume for the corresponding holidays in the previous four years, and obtain the first predicted ticket sales volume for the holiday based on the comprehensive ticket score data for the corresponding holiday in the previous year; The second prediction module predicts the ticket sales of each type of ticket on the next day based on the average ticket sales data of each type of ticket in the previous four weeks and the average ticket sales data under various weather conditions, and obtains the second predicted ticket sales Fts of each type of ticket on the next day. i If the next day is a holiday, the average ticket sales of each type of ticket during the holiday in the past three years is combined with the weather forecast data of the next day to predict the ticket sales of each type of ticket on the next day; Comprehensive forecasting module, based on the actual ticket sales data of various types of tickets per day in the previous week Ats io , and the first predicted ticket sales volume Yts for each type of ticket io and the second predicted ticket sales Ets io , respectively calculate the first error Fe of the first predicted ticket sales volume and the second predicted ticket sales volume relative to the actual ticket sales volume data io and the second error Se io , according to the first error Fe io and the second error Se io Get the weight factors β1 and β2 of the first predicted ticket sales and the second predicted ticket sales; convert the first predicted ticket sales Yts of each type of ticket for the next day into io and the second predicted ticket sales Ets io Combined with the corresponding weight factor, the comprehensive forecast ticket sales volume Cts of each type of ticket for the next day is obtained i ; The ticket channel adjustment module simulates the network status of each type of ticket when different numbers of tickets are issued in a single channel, and obtains the single channel ticket threshold St for each type of ticket. i ; Based on the comprehensive forecast of ticket sales volume Cts for each type of ticket i The single-channel ticket volume threshold St i The relationship between the number of ticketing channels for each type of ticket is adjusted; when a new ticketing request appears in the ticketing waiting queue, it is determined whether the number of ticketing channels needs to be adjusted.

2. The intelligent ticketing system according to claim 1, characterized in that: At the end of each day, the ticket sales data of each type of ticket in the same week as the next day in the previous four weeks are obtained from the ticket database, recorded as Pw ij , where j represents the previous jth week; obtain the weather forecast data for the next day, and use We to represent the weather forecast data for the next day. When We=1, it means it is a rainy or snowy day, and when We=2, it means it is a non-rainy or snowy day.

3. An intelligent ticketing system as claimed in claim 2, characterized in that: Count the actual number of ticket purchases by all users in the smart ticketing system k , where k represents the kth user, and the actual number of ticket purchases represents the number of times a user purchases tickets normally and does not refund them; according to the actual number of ticket purchases Atp by all users in the smart ticketing system k , calculate the rating reliability index Sri for each user k , the expression is: Among them, Mi k is the member index, and Mi k ∈{1,N}, and N>1, when the user is a non-member user, Mi k =1; when the user is a member user, Mi k =N; the value of Max{x} is the maximum value of x; Get the ticket sales data of various types of tickets in the previous week, and get the names of the users who purchased the tickets of various types of tickets. According to the names of the users who purchased the tickets, get the user's rating data for each type of ticket in the corresponding user table. ik and each user's rating reliability index Sri k Combined to get the comprehensive score Os of various types of tickets in the previous week i , the expression is: Among them, m i Represents the total number of tickets actually purchased for the i-th type of ticket.

4. An intelligent ticketing system as claimed in claim 3, characterized in that: Obtain the comprehensive score data of each type of ticket in the second week, the third week and the fourth week from the weekly record table in the ticket database, as well as the actual ticket sales data of each type of ticket in the first week, the second week and the third week; based on the linear regression equation Y ij =b+a*X i(j+1) The comprehensive score data of each type of ticket in the second week, the third week, and the fourth week, as well as the actual ticket sales data of each type of ticket in the first week, the second week, and the third week are fitted, where Y ij represents the actual sales volume of the ith type of ticket in the previous j week, X i(j+1) represents the comprehensive score of the i-th type of ticket in the previous j+1 week; the slope a0 and intercept b0 of the fitting line are calculated by the least squares method, and the fitting equation is Y ij =b0+a0*X i(j+1) .

5. An intelligent ticketing system as claimed in claim 4, characterized in that: The comprehensive score data of various types of tickets in the previous week, that is, the previous week, Os i Substituting into the fitting equation, we can get the first predicted ticket sales volume Fts for each type of ticket in the next week. i ; Filter out the ticket sales of each type of ticket on weekdays and weekends in the first week, the second week, the third week, and the fourth week, and calculate the ticket sales of each type of ticket on weekdays Wts in each week ij And the weekend ticket sales volume Sts ij , through the formula Calculate the average ratio of ticket sales on weekdays to weekends for each type of ticketing i , where n represents the number of weeks without holidays in the first week, the second week, the third week, and the fourth week, and n∈{1,…,4}; the first predicted ticket sales volume Fts of each type of ticket in the next week i The average ratio of ticket sales on weekdays to weekends for all types of ticket sales i Combined, we get the first predicted total ticket sales volume Wts for each type of ticket on the next weekday i , and the first forecast total ticket sales volume of various types of tickets on weekends Sts i ; Then get the first predicted ticket sales volume Yts for each day of the next week i , the first predicted ticket sales volume for each type of ticket for the next weekday For the weekends in the next week, the first forecast ticket sales of each type of ticket For holidays within the next week, obtain the name of the holiday and obtain the ticket sales data of each type of ticket within the previous year, the previous year, the previous year, the previous year, and the previous year from the ticket database, and obtain the names of the users who purchased each type of ticket in each year, obtain the user's rating data for each type of ticket in the corresponding user table according to the name of the ticket purchasing user, and calculate the comprehensive ratings of each type of ticket within the previous year, the previous year, the previous year, and the previous year through the comprehensive rating calculation formula; combine the comprehensive ratings of each type of ticket within the previous year, the previous year, and the previous year with the actual ticket sales data of each type of ticket within the previous year, the previous year, and the previous year, respectively, to construct a linear fitting function for the corresponding holiday, and bring the comprehensive ratings of each type of ticket within the previous year into the linear fitting function to obtain the first predicted ticket sales volume Yts for the holiday in the next week. i .

6. An intelligent ticketing system as claimed in claim 5, characterized in that: At the end of each day, the ticket sales data of each type of ticket for the previous four weeks are obtained from the ticket database, and the ticket sales data of each type of ticket for the same week in the previous four weeks are combined to obtain the average ticket sales data of each type of ticket for each day XAts i(week) , where week=1 means Monday, and so on; Get the ticket sales data for each type of weather from the ticket database, combine the ticket sales data for each type of weather, and get the average ticket sales data for each type of ticket under each type of weather. i(we) ; Get the next day's week information and weather forecast data; If the next day is not a holiday, the second predicted ticket sales volume Ets for each type of ticket on the next day is obtained by combining the average ticket sales volume data of each type of ticket on the corresponding week of the next day and the average ticket sales volume data of each type of ticket on the corresponding weather. i , the expression is: You are i =α1*XAts i(week) +α2*WAts i(we) Among them, α1 and α2 are the weight coefficients of week information and weather information respectively.

7. An intelligent ticketing system as claimed in claim 6, characterized in that: If the next day is a holiday, obtain the name of the holiday and obtain the ticket sales data of each type of ticket on the holiday in the past three years from the ticket database. Combine the three obtained ticket sales data of each type of ticket on the holiday to obtain the average ticket sales of each type of ticket on the holiday. i(ho) , where ho represents different holidays; the average ticket sales of each type of ticket on the next day corresponding to the holiday is combined with the average ticket sales of each type of ticket on the corresponding weather to obtain the second predicted ticket sales of each type of ticket on the next day Ets i , the expression is: Ets i =α1*HAts i(ho) +α2*WAts i(we) .

8. An intelligent ticketing system as claimed in claim 7, characterized in that: Get the actual ticket sales data of each type of ticket every day in the previous week from the ticket database Ats io , and the first predicted ticket sales volume Yts for each type of ticket every day in the previous week io and the second predicted ticket sales Ets io , where o represents the oth day in the previous week; the actual ticket sales data Ats io and the first predicted ticket sales volume Yts io and the second predicted ticket sales Ets io After combining, we get the first error Fe between the first predicted ticket sales volume and the second predicted ticket sales volume relative to the actual ticket sales volume data. io and the second error Se io , the expression is Fe io =|Ats io -Yts io |, Se io =|Ats io -Ets io |; The first daily error Fe in the previous week io and the second error Se io Take the average and get the first average error and the second mean error The first average error The second mean error Combined, the weight factors β1 and β2 of the first predicted ticket sales volume and the second predicted ticket sales volume are calculated respectively, and the expressions are: Combine the first predicted ticket sales volume and the second predicted ticket sales volume of each type of ticket for the next day with the corresponding weight factor to obtain the comprehensive predicted ticket sales volume Cts of each type of ticket for the next day i , the expression is: Cts i =β1*Yts i +β2*Ets i .

9. An intelligent ticketing system as claimed in claim 8, characterized in that: A simplified model of the intelligent ticketing system is built. In the simplified model, the ticketing requests corresponding to each type of ticketing are simulated to process the ticketing through a single channel. Ticketing requests are sent to the simplified model of the intelligent ticketing system one by one, and the network status data of the intelligent ticketing system is monitored in real time. The real-time network status data is analyzed through the network analysis tool in the system. When the network status is first displayed as congested, the total number of ticketing requests at this time is recorded. The total number of ticketing requests obtained by simulating each type of ticketing when the network status is first displayed as congested is used as the single-channel ticketing threshold St for the corresponding type of ticketing. i .

10. An intelligent ticketing system as claimed in claim 9, characterized in that: The comprehensive forecast ticket sales volume Cts of each type of ticket i The single-channel ticket volume threshold St i Combined, if Cts i <St i , two channels are allocated for this type of ticket to process the ticket; if Cts i ≥St i , then set it as the ticket allocation of this type There are three channels for ticket issuance: Indicates that x is rounded down; Whenever a new ticket waiting request appears in the ticket waiting queue of each channel, the total number of ticket waiting requests Npr in the ticket waiting queue of the channel is counted. i , if Npr i ≥0.9*St i , you need to use a new channel for ticketing operations. Specifically, take out the ticketing waiting requests sorted as even numbers in the ticketing waiting queue of the current channel, put them into the ticketing waiting queue of a new channel, and perform subsequent ticketing processing operations.

Images