CN-121981543-A - Comprehensive management method and system for automobile passenger ticket business and operation

CN121981543ACN 121981543 ACN121981543 ACN 121981543ACN-121981543-A

Abstract

The application provides a comprehensive management method and a comprehensive management system for automobile passenger ticket business and operation, which relate to the technical field of automobile passenger ticket business management, wherein the method comprises the steps of collecting historical operation data and target date characteristics of a target line; the method comprises the steps of constructing a passenger number prediction model based on historical operation data, inputting a target date characteristic into the passenger number prediction model, outputting a predicted passenger flow result of a target line in a target date, evaluating the full-load risk of vehicles at station points on each running water number in the target date according to the predicted passenger flow result, and displaying a risk evaluation result after a user purchases a ticket in the target date of the target line. By introducing the dynamic full-load risk assessment and real-time passenger flow prediction model, the passenger flow of different time periods and stations can be estimated and the full-load risk assessment and display can be carried out, so that the running efficiency of the inter-city and urban and rural coach bus is improved, a user can select a proper trip period, and the riding experience of the user is optimized.

Inventors

YAN JIAMIN
JIN HAO
LIU YUSONG

Assignees

山东悦程网络科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. An integrated management method for passenger ticket business and operation of an automobile, which is characterized by comprising the following steps: collecting historical operation data and target date characteristics of a target line, wherein the historical operation data comprises historical date characteristics, ticket selling data in corresponding departure time periods and the number of boarding persons at each boarding station in each running train number; The passenger flow prediction comprises the steps of constructing a passenger number prediction model based on historical operation data, inputting a target date characteristic into the passenger number prediction model, and outputting a predicted passenger flow result of a target line in a target date, wherein the predicted passenger flow result comprises predicted ticket selling data in each departure time period and predicted boarding number of each boarding station point in each running water passenger number; the risk assessment is carried out on the full-load risk of the vehicles at the station points on each running water train in the target date according to the predicted passenger flow result, and the full-load risk level of the stations on each running water train in each running water train is obtained; And displaying the full-load risk level of each boarding station in each running water train number in the target date after the user purchases the ticket in the target date of the target line.
2. The integrated management method for passenger ticket business and operation of automobile according to claim 1, wherein in the step of passenger flow prediction, a passenger number prediction model is constructed based on historical operation data, comprising: Constructing a model, namely constructing a passenger number prediction model, wherein the model comprises an input layer, a shared characteristic layer, a constraint perception layer and a multi-task output layer; The feature extraction, namely constructing a training set based on historical operation data, inputting the training set into a passenger number prediction model, extracting shared features of the historical operation data by utilizing a shared feature layer, and creating task specific branches based on ticket selling data in each departure time period and the number of passengers on each boarding station point in each running number, wherein the task specific branches comprise ticket selling data tasks and site boarding number tasks; rule constraint, namely inputting the shared features into a constraint perception layer, and performing constraint perception modeling on the shared features based on time sequence association relations between departure time periods, driving sequence relations of boarding stations and vehicle passenger capacity constraint to generate constraint features meeting passenger traffic rules; And model optimization, namely defining a multi-task loss function, inputting the shared feature into a task specific branch, performing task processing based on constraint features, calculating a multi-task loss value in the task processing process, iteratively optimizing a passenger number prediction model by minimizing the task loss function, and recording the optimized passenger number prediction model as a new passenger number prediction model.
3. The integrated management method for passenger ticket business and operation according to claim 2, wherein the passenger number prediction model further comprises a deviation correction layer; after the step of performing data acquisition, before the step of performing passenger flow prediction, further comprising: acquiring real-time data in each departure time period of departure in a target date of a target line, wherein the real-time data comprises actual ticketing data in each departure time period and actual number of vehicles at each station; Inputting the target date characteristic into a passenger number prediction model, and outputting a predicted passenger flow result of the target line in the target date, wherein the method comprises the following steps: Inputting the characteristics of a target date into a passenger number prediction model, and outputting an initial predicted passenger flow result of a target line in the target date, wherein the initial predicted passenger flow result comprises initial predicted ticket selling data in each departure time period and initial predicted passenger numbers of all boarding stations; Calculating the difference between the real-time data in each departure time period of departure in the target date and the initial predicted passenger flow result in the corresponding departure time period to obtain the data deviation of each departure time period, wherein the data deviation comprises ticket selling data deviation and boarding number deviation; and (3) deviation correction, namely constructing deviation features based on the obtained data deviation and the time position information of the corresponding departure time period, and inputting the deviation features and the initial predicted passenger flow result into a deviation correction layer to correct the initial predicted passenger flow result in each departure time period which does not occur.
4. A method of integrated management of automotive passenger ticketing and operation as set forth in claim 3, wherein after the step of calculating the deviation is performed, before the step of performing the deviation correction, further comprising: And (3) correction judgment, namely judging whether the preset deviation correction rule is met or not based on the data deviation of each departure time period: If yes, a deviation correction step is carried out on the initial predicted passenger flow result, and then the corrected initial predicted passenger flow result is used as a predicted passenger flow result; if not, taking the initial predicted passenger flow result as a predicted passenger flow result.
5. The integrated management method for automotive passenger ticket business and operation according to claim 4, wherein the deviation correction rule includes a threshold judgment rule and a homodromous judgment rule; Threshold judgment rules, namely in each departure time period, ticket selling data deviation or boarding number deviation exceeds a corresponding preset data threshold; the same direction judging rule is that the directions of ticket selling data deviation or boarding number deviation in all the occurring departure time periods are consistent; when ticket selling data deviation or boarding number deviation of each departure time period meets any one of the threshold judgment rule and the same direction judgment rule, the data deviation of each departure time period is judged to be in accordance with the deviation correction rule.
6. A method of integrated management of automotive passenger ticketing and operation as set forth in claim 3, wherein said step of bias correction includes: Building deviation characteristics: ; bias modeling, namely constructing a time distance function and a propagation weight function; the calculation formula of the time distance function is as follows ; The propagation weight function is calculated as ; Correcting deviation: the formula for correcting the number of passengers is as follows: ; the formula for correcting ticketing data is as follows: ; Wherein, the Indicating the deviation of the ticketing data, , Indicating a departure time period In the actual ticketing data of the ticket, Indicating a departure time period In the initial predictive ticketing data of the interior, The deviation of the number of people getting on the car is indicated, , Indicating a departure time period Internal boarding station point In fact the number of people in the vehicle, Indicating a departure time period Internal boarding station point Is used for the initial prediction of the number of people on the vehicle, Indicating that the departure time period has elapsed, , , Indicating that a departure period has not occurred, Representing departure time period At a time-sequential location within the target date, Departure time period At a time-sequential location within the target date, Representing departure time period With departure time period Is used for the time distance of (a), The time-decay factor is represented as such, , Represent the first A departure time period has elapsed and the departure time period, Representing corrected departure time period Internal boarding station point Is used for the initial prediction of the number of people on the vehicle, Indicating a departure time period Internal boarding station point Is used for the initial prediction of the number of people on the vehicle, Representing corrected departure time period In the initial predictive ticketing data of the interior, Indicating a departure time period Initial predictive ticketing data within.
7. The method for integrated management of automotive passenger ticketing and operation of claim 1, wherein the step of risk assessment comprises: Calculating the predicted passenger number, namely accumulating the predicted passenger number of each boarding station point according to the sequence of each boarding station point based on the predicted passenger number of each boarding station point in each running water train number in each departure time period; The full-load risk coefficient is calculated, namely the preset passenger carrying number is obtained, and the full-load risk coefficient of each boarding station is obtained based on the ratio of the predicted passenger carrying number and the preset passenger carrying number of each boarding station; And (3) risk level division, namely obtaining the full-load risk level of each upper station point based on a preset full-load risk coefficient level division table and the full-load risk coefficients of each upper station point.
8. The integrated management method for automotive passenger ticketing and operation according to claim 1, characterized by further comprising, after the step of performing risk assessment, before the step of performing risk level presentation: Time period sequencing, namely acquiring a boarding station which a user needs to ride, sequencing the full-load risk level of the boarding station in each departure time period according to the boarding station selected by the user from small to large to obtain a risk sequencing result of the boarding station; and in the step of risk level display, displaying the risk ranking result of the boarding station to the user.
9. The integrated management method for passenger ticket business and operation of claim 1, wherein in the step of displaying the risk level, the number of browsing of the user for each boarding station in each departure time period is obtained, and the risk level of each boarding station in each departure time period is corrected based on the number of browsing of the user.
10. An integrated management system for automotive passenger ticketing and operation, characterized in that it is adapted to a system according to any one of claims 1-9, said system comprising: the data acquisition module is used for acquiring historical operation data and target date characteristics of the target line, wherein the historical operation data comprises the historical date characteristics, ticket selling data in corresponding departure time periods and the number of boarding persons at each boarding station point in each running train number; the passenger flow prediction module is used for constructing a passenger number prediction model based on historical operation data, inputting a target date characteristic into the passenger number prediction model, and outputting a predicted passenger flow result of a target line in a target date, wherein the predicted passenger flow result comprises predicted ticket selling data in each departure time period and predicted boarding number of each boarding station point in each running water number; The risk assessment module is used for assessing the full-load risk of the vehicles at the station points on each running water train in the target date according to the predicted passenger flow result to obtain the full-load risk level of the stations on each running water train; And the risk level display module is used for displaying the full-load risk level of each boarding station in each running water train number in the target date after the user purchases the ticket in the target date of the target line.

Description

Comprehensive management method and system for automobile passenger ticket business and operation Technical Field The application relates to the technical field of automobile passenger ticket business management, in particular to an automobile passenger ticket business and operation integrated management method and system. Background With the rapid development of inter-city traffic and urban and rural traffic, long-distance buses still play an important role in public transportation in the scenes of inter-city travel, urban and rural engagement and the like in medium and short distances. Compared with the transportation modes of railways, aviation and the like which need to lock shifts in advance, the coach has the characteristics of flexible departure, scattered stations, capability of getting on and off passengers on the way and the like, and a running water departure or quasi-running water departure mode is widely adopted to adapt to passenger flow changes in different time periods. In the inter-city and urban and rural long distance bus operation, an operation mode of 'fixed line-running number of vehicles-multiple boarding stations' is commonly adopted, namely, the same line continuously launches in a certain time interval in one day, the vehicles are allowed to get on a plurality of intermediate stations besides the starting stations, the number of the passengers on a single vehicle is fixed, and the continuous getting on the vehicle is not allowed after the vehicle is fully loaded. Because the travel time selection of the passengers has obvious concentration, the passenger flow distribution of different time periods and different boarding stations has obvious difference, so that vehicles in partial periods are frequently fully loaded, and the empty rate of vehicles in other periods is higher. Especially in the scene of supporting the boarding of the midway website, the quantity of the rest seats when the vehicle arrives at the midway website has larger uncertainty, and the problems that passengers cannot get in bus and travel experience is poor and the like after waiting for the station are easy to occur because the existing coach management method lacks prediction and prompt mechanisms for different time periods and the full-load risk of the website, so that the passengers cannot know the crowded degree information of each time period before buying tickets or traveling, and only experience or randomly select the riding time. Disclosure of Invention In order to overcome the defects, the application provides a comprehensive management method and system for automobile passenger ticket business and operation. In a first aspect, the present application provides a method for comprehensive management of passenger ticket service and operation of an automobile, which adopts the following technical scheme: a method for integrated management of automotive passenger ticketing and operation, the method comprising: collecting historical operation data and target date characteristics of a target line, wherein the historical operation data comprises historical date characteristics, ticket selling data in corresponding departure time periods and the number of boarding persons at each boarding station in each running train number; The passenger flow prediction comprises the steps of constructing a passenger number prediction model based on historical operation data, inputting a target date characteristic into the passenger number prediction model, and outputting a predicted passenger flow result of a target line in a target date, wherein the predicted passenger flow result comprises predicted ticket selling data in each departure time period and predicted boarding number of each boarding station point in each running water passenger number; the risk assessment is carried out on the full-load risk of the vehicles at the station points on each running water train in the target date according to the predicted passenger flow result, and the full-load risk level of the stations on each running water train in each running water train is obtained; And displaying the full-load risk level of each boarding station in each running water train number in the target date after the user purchases the ticket in the target date of the target line. Optionally, in the step of passenger flow prediction, constructing a passenger number prediction model based on the historical operation data includes: Constructing a model, namely constructing a passenger number prediction model, wherein the model comprises an input layer, a shared characteristic layer, a constraint perception layer and a multi-task output layer; The feature extraction, namely constructing a training set based on historical operation data, inputting the training set into a passenger number prediction model, extracting shared features of the historical operation data by utilizing a shared feature layer, and creating task specific branches based on ticket selling data in eac