A method for generating single-line bus passenger travel data

Abstract

The present invention discloses a method for generating single-line bus passenger travel data, first randomly obtains G candidate routes, then obtains the spatial vectors of the sites of all candidate routes based on the site spatial information, then obtains the total spatial similarity values ​​of all candidate routes and the target route based on the spatial vectors, and selects the candidate routes corresponding to the first Q total spatial similarity values ​​as preferred candidate routes, then obtains the similarity between each preferred candidate route and the target route by calculating the data similarity index and similarity, selects the preferred candidate route with the highest similarity as the optimal candidate route, and finally uses the optimal candidate route as a learning sample to generate passenger travel data of the target sample through a cyclic generative adversarial network algorithm. The present invention can generate a large amount of bus passenger travel data of the target route based on real data, conveniently, quickly and at low cost, and the generated data conforms to the real law and can be applied to fields such as bus analysis.

Description

Technical Field

[0001] The present invention relates to the field of transportation engineering, and belongs to the category of urban public transportation. More specifically, it relates to a method for generating travel data of single-line public transportation passengers. Background Art

[0002] Public transportation passenger travel data refers to a series of data records generated by passengers using public transportation, including passenger boarding station information, alighting station information, corresponding boarding and alighting time, and line number, etc. Public transportation passenger travel data can be used to analyze passenger flow and classify passengers in urban public transportation systems. Public transportation travel information is widely used in urban public transportation operation and scheduling.

[0003] With the rapid development of mobile Internet, a series of innovations have been made in the payment methods of public transportation. The payment methods adopted by urban public transportation in my country are segmented billing system and one-ticket system. The former calculates the fee by recording the passenger boarding and disembarking information, while the latter only records the passenger boarding time and space information but not the disembarking time and space information. For urban public transportation that adopts the one-ticket system, there is a lack of passenger disembarking information data, and the public transportation passenger travel data is incomplete. At present, the commonly used means are to collect data through the following method and the disembarkation site inference method to complete the data, but the disembarkation inference method relies on data assumptions and has low accuracy, while the following method has a complicated and cumbersome data process and relies on manual tabulation and recording, which is inefficient and costly. Summary of the invention

[0004] The present invention provides a method for generating travel data of single-line bus passengers. The present invention can generate travel data of single-line target bus passengers under the bus one-ticket system conveniently, quickly and at low cost based on an existing complete travel data set of similar single-line buses, thereby providing necessary data conditions for application scenarios such as bus analysis.

[0005] In order to achieve the above object, the technical solution of the present invention is:

[0006] A method for generating single-line bus passenger travel data comprises the following steps:

[0007] S1. Randomly select G bus routes as candidate routes, and obtain the historical riding data and station space information of all candidate routes;

[0008] S2. Determine the spatial vectors of all the sites in the G candidate routes according to the site spatial information;

[0009] S3. Calculate the spatial similarity index between each station in the target line and all stations in any candidate line according to the spatial vector of the station, and obtain the total spatial similarity value between the candidate line and the target line;

[0010] S4. For each candidate route, a total spatial similarity value with the target route is obtained in the manner of step S3. Sort the total spatial similarity values of all candidate routes and the target route from largest to smallest, and select the top Q candidate routes corresponding to the total spatial similarity values as the preferred candidate routes;

[0011] S5. Calculate the data similarity index between each station in the target route and all stations in any one of the preferred candidate routes to obtain the total data similarity value between this preferred candidate route and the target route;

[0012] S6. For each preferred candidate route, a total data similarity value with the target route is obtained in the manner of step S5. According to the obtained total data similarity value, calculate the similarity between each preferred candidate route and the target route, and select the preferred candidate route corresponding to the maximum similarity value as the optimal candidate route;

[0013] S7. Use the optimal candidate route and the target route as the learning sample and the target sample;

[0014] S8. Through the cyclic generative adversarial network algorithm, generate the passenger travel data of the target sample from the travel data of the learning sample.

[0015] The method of the present invention first randomly obtains G bus routes as candidate routes, then according to the station spatial information, obtains the spatial vectors of all stations of all candidate routes, and then according to the spatial vectors, obtains the total spatial similarity values of all candidate routes and a pre-selected target route, and selects the top Q (Q < G, both are positive integers) candidate routes corresponding to the total spatial similarity values as the preferred candidate routes. Then, from these Q preferred candidate routes, by calculating the data similarity index and similarity, the similarity between each preferred candidate route and the target route is obtained. After that, the preferred candidate route with the highest similarity is selected as the optimal candidate route. Finally, using this optimal candidate route as the learning sample and the target route as the target sample, the travel data of the learning sample can generate the passenger travel data of the target sample through the cyclic generative adversarial network algorithm.

[0016] Further, in step S2, the station spatial information includes longitude and latitude information and point-of-interest classification information. The process of determining the station spatial vector of the candidate route according to the station spatial information is as follows:

[0017] Determine the amount of point-of-interest data under the longitude, latitude and tolerance radius of each station. Denote the longitude and latitude information of each station as a 2-dimensional vector l, and the tolerance radius of the station as r. Count the number of k types of points of interest within the tolerance radius r of the station, denoted as a k-dimensional vector h. Then the spatial vector of each station is represented by a (2 + k)-dimensional vector (l, h).

[0018] Further, in step S3, the calculation formula of the spatial similarity index is as follows:

[0019]

[0020] In the formula, s i is the i-th station in the target line, b j is the jth station in the candidate line, P(s i , b j ) is site s i and site b j The spatial similarity index, l i 、h i For sites i The latitude and longitude and point of interest information, j 、h j For site b j The longitude and latitude and interest point information, α and β are the influence coefficients of the longitude and latitude and interest point information on the spatial similarity index. The value range of the spatial similarity index P is between [-1, 1]. The larger the value, the higher the spatial similarity of the two sites.

[0021] Furthermore, in step S3, the specific process of obtaining the total spatial similarity value between the candidate route and the target route is as follows:

[0022] Assume that the target line has n stations, that is, s i ∈{s1,s2,...,s n}, i = 1, 2, ..., n, the candidate line has m stations, that is, b j ∈{b1,b2,...,b m}, j = 1, 2, ..., m;

[0023] First, the spatial similarity index of the site s1 in the target line and the m sites in the candidate line is calculated to obtain the m spatial similarity indexes corresponding to the site s1 and the m sites in the candidate line. The maximum value is selected as the maximum value P1 of the spatial similarity index between the site s1 and the candidate line. Similarly, the set of the maximum values ​​of the spatial similarity indexes corresponding to the n sites in the target line and the candidate line is obtained, which is recorded as {P1, P2, ..., P n}, and finally P1, P2, ..., P n The sum is accumulated to obtain the total spatial similarity value between the candidate route and the target route.

[0024] Further, in step S5, the calculation formula of the data similarity index is as follows:

[0025]

[0026] In the formula, E(s i , b j ) is site s iand site b j The data similarity index, R s , R b are the total number of passengers on the target route and the preferred candidate route, T u , T d are the stations of the target line respectively i The number of people getting on and off the bus, C u , C d Select the candidate line stations b respectively j The data similarity index E ranges from [0, 1]. The larger the value, the lower the data similarity between the two stations.

[0027] Further, in step S5, the specific process of obtaining the total data similarity value between the preferred candidate route and the target route is as follows:

[0028] First, the data similarity index of the site s1 in the target line and the m sites in the preferred candidate line is calculated to obtain the m data similarity indexes corresponding to the site s1 and the m sites in the preferred candidate line. The maximum value is selected as the maximum value E1 of the data similarity index corresponding to the site s1 and the preferred candidate line. Similarly, the set of the maximum values ​​of the data similarity indexes corresponding to the n sites in the target line and the preferred candidate line is obtained, which is recorded as {E1, E2, ..., E n}, and finally E1, E2, ..., E n The data are accumulated to obtain the total value of data similarity between the preferred candidate route and the target route.

[0029] Further, in step S6, the similarity calculation formula is as follows:

[0030]

[0031] In the formula, V(S,b) is the similarity between the preferred candidate route and the target route, E i ∈{E1,E2,…,E n}, E i is the site s in the target line i The maximum value of the data similarity index corresponding to the preferred candidate line, that is, is the total data similarity value between the preferred candidate route and the target route. The similarity V ranges from [0,1]. The larger the value, the more similar the preferred candidate route is to the target route.

[0032] Furthermore, in step S7, the one-ticket riding dataset of the target route is used as the data source of the target sample, and the complete passenger riding dataset of the optimal candidate route is used as the data source of the learning sample.

[0033] Furthermore, in step S8, a cyclic generative adversarial network model is used as a generator, the travel data matrix of the learning sample is input into the generator, simulation data is iteratively generated in the generator, and the generated simulation data is put into the judge for judgment. If the similarity of the simulation data is not lower than the set threshold, the generated simulation data is considered valid and output as the passenger travel data of the target sample.

[0034] Furthermore, in the decision maker, the similarity between the simulated data and the real travel data of the learning sample is first calculated, and then the similarity is compared with the set threshold. If the similarity of the simulated data is not lower than the set threshold, the generated simulated data is considered valid and output;

[0035] The set threshold refers to the similarity between the optimal candidate route obtained in step S6 and the target route.

[0036] The beneficial effects of the present invention are:

[0037] The method for generating bus passenger travel data provided by the present invention can generate passenger travel data of a target route according to the spatial information of bus stops and the real data of the optimal candidate route. Compared with the traditional bus following method and the alighting stop inference method, the method can generate bus passenger travel data of a target route under the bus one-ticket system conveniently, quickly and at low cost based on the historical real data of the optimal candidate route. Only the spatial information of the stop is required. The method for generating data has high applicability and the obtained data has high accuracy, and can be applied to fields such as bus analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of the method for generating single-line bus passenger travel data of the present invention. DETAILED DESCRIPTION

[0039] The drawings are only for illustrative purposes and cannot be construed as limiting the present invention. To better illustrate the present embodiment, some parts of the drawings may be omitted, enlarged, or reduced, and do not represent the size of the actual product. For those skilled in the art, it is understandable that some well-known structures and their descriptions may be omitted in the drawings. The positional relationships described in the drawings are only for illustrative purposes and cannot be construed as limiting the present invention.

[0040] Embodiment 1:

[0041] like Figure 1 As shown, a method for generating single-line bus passenger travel data includes the following steps:

[0042] S1. Randomly select G bus routes as candidate routes, and obtain the historical riding data and station space information of all candidate routes;

[0043] S2. Determine the spatial vectors of all stations in the G candidate lines according to the station spatial information;

[0044] S3. According to the spatial vectors of the stations, calculate the spatial similarity index between each station in the target line and all stations in any one of the candidate lines, and obtain the total spatial similarity value between this candidate line and the target line;

[0045] S4. Each candidate line obtains a total spatial similarity value with the target line in the manner of step S3. Sort the total spatial similarity values of all candidate lines and the target line from large to small, and select the candidate lines corresponding to the top Q total spatial similarity values as the preferred candidate lines;

[0046] S5. Calculate the data similarity index between each station in the target line and all stations in any one of the preferred candidate lines, and obtain the total data similarity value between this preferred candidate line and the target line;

[0047] S6. Each preferred candidate line obtains a total data similarity value with the target line in the manner of step S5. According to the obtained total data similarity value, calculate the similarity between each preferred candidate line and the target line, and select the preferred candidate line corresponding to the maximum similarity value as the optimal candidate line;

[0048] S7. Use the optimal candidate line and the target line as the learning sample and the target sample;

[0049] S8. Through the cyclic generative adversarial network algorithm, generate the passenger travel data of the target sample from the travel data of the learning sample.

[0050] The method of the present invention first randomly obtains G bus lines as candidate lines, and then according to the station spatial information, obtains the spatial vectors of all stations of all candidate lines. Then, according to the spatial vectors, obtains the total spatial similarity values between all candidate lines and a pre-selected target line, and selects the candidate lines corresponding to the top Q (Q < G, both are positive integers) total spatial similarity values as the preferred candidate lines. Then, through the calculation of the data similarity index and similarity from these Q preferred candidate lines, obtains the similarity between each preferred candidate line and the target line. Then, selects the preferred candidate line with the highest similarity as the optimal candidate line. Finally, uses this optimal candidate line as the learning sample and the target line as the target sample, and the travel data of the learning sample can generate the passenger travel data of the target sample through the cyclic generative adversarial network algorithm.

[0051] In this embodiment, the values of G and Q can be set according to the actual situation. For example, 100 bus lines can be randomly selected as candidate lines, that is, G = 100. After obtaining the total spatial similarity values between all candidate lines and a pre-selected target line, select the candidate lines corresponding to the top 10 total spatial similarity values as the preferred candidate lines, that is, Q = 10.

[0052] In step S2 of this embodiment, the site space information includes latitude and longitude information and interest point classification information. The process of determining the site space vector of the candidate route according to the site space information is as follows:

[0053] Determine the amount of POI data under the longitude and latitude of each station and the tolerance radius, record the longitude and latitude information of each station as a 2-dimensional vector l, record the station's tolerable radius as r, and count the number of k types of POIs under the tolerable radius r of the station, recorded as a k-dimensional vector h. Then the spatial vector of each station is represented by a (2+k)-dimensional vector (l, h).

[0054] In step S3 of this embodiment, the calculation formula of the spatial similarity index is as follows:

[0055]

[0056] In the formula, s i is the i-th station in the target line, b j is the jth station in the candidate line, P(s i , b j ) is site s i and site b j The spatial similarity index, l i 、h i For sites i The latitude and longitude and point of interest information, j 、h j For site b j The longitude and latitude and interest point information, α and β are the influence coefficients of the longitude and latitude and interest point information on the spatial similarity index. The value range of the spatial similarity index P is between [-1, 1]. The larger the value, the higher the spatial similarity of the two sites.

[0057] According to the spatial similarity index, the specific process of obtaining the total spatial similarity value between the candidate route and the target route is as follows:

[0058] Assume that the target line has n stations, that is, s i ∈{s1,s2,...,s n}, i = 1, 2, ..., n, the candidate line has m stations, that is, b j ∈{b1,b2,...,b m}, j = 1, 2, ..., m;

[0059] First, the spatial similarity index of the site s1 in the target line and the m sites in the candidate line is calculated to obtain the m spatial similarity indexes corresponding to the site s1 and the m sites in the candidate line. The maximum value is selected as the maximum value P1 of the spatial similarity index between the site s1 and the candidate line. Similarly, the set of the maximum values ​​of the spatial similarity indexes corresponding to the n sites in the target line and the candidate line is obtained, which is recorded as {P1, P2, ..., P n}, and finally P1, P2, ..., P n The sum is accumulated to obtain the total spatial similarity value between the candidate route and the target route.

[0060] In step S5 of this embodiment, the calculation formula of the data similarity index is as follows:

[0061]

[0062] In the formula, E(s i , b j ) is site s i and site b j The data similarity index, R s , R b are the total number of passengers on the target route and the preferred candidate route, T u , T d are the stations of the target line respectively i The number of people getting on and off the bus, C u , C d Select the candidate line stations b respectively j The data similarity index E ranges from [0, 1]. The larger the value, the lower the data similarity between the two stations.

[0063] According to the data similarity index, the specific process of obtaining the total data similarity value between the preferred candidate route and the target route is as follows:

[0064] First, the data similarity index of the site s1 in the target line and the m sites in the preferred candidate line is calculated to obtain the m data similarity indexes corresponding to the site s1 and the m sites in the preferred candidate line. The maximum value is selected as the maximum value E1 of the data similarity index corresponding to the site s1 and the preferred candidate line. Similarly, the set of the maximum values ​​of the data similarity indexes corresponding to the n sites in the target line and the preferred candidate line is obtained, which is recorded as {E1, E2, ..., E n}, and finally E1, E2, ..., E n The data are accumulated to obtain the total value of data similarity between the preferred candidate route and the target route.

[0065] In step S6 of this embodiment, the similarity calculation formula is as follows:

[0066]

[0067] In the formula, V(S, b) is the similarity between the preferred candidate route and the target route, E i ∈{E1,E2,...,E n}, E i is the site s in the target line i The maximum value of the data similarity index corresponding to the preferred candidate line, that is, is the total data similarity value between the preferred candidate route and the target route. The similarity V ranges from [0, 1]. The larger the value, the more similar the preferred candidate route is to the target route.

[0068] In step S7 of this embodiment, the one-ticket riding dataset of the target route is used as the data source of the target samples, and the complete passenger riding dataset of the optimal candidate route is used as the data source of the learning samples.

[0069] In step S8 of this embodiment, a cyclic generative adversarial network model is used as a generator, the travel data matrix of the learning sample is input into the generator, simulation data is iteratively generated in the generator, and the generated simulation data is put into the judge for judgment. If the similarity of the simulation data is not lower than the set threshold, the generated simulation data is considered valid and output as the passenger travel data of the target sample.

[0070] Among them, in the decision device, the similarity between the simulated data and the real travel data of the learning sample is first calculated, and then the similarity is compared with the set threshold. If the similarity of the simulated data is not lower than the set threshold, the generated simulated data is considered valid and output. The set threshold refers to the similarity between the optimal candidate route obtained in step S6 and the target route.

[0071] The method for generating bus passenger travel data provided by the present invention can generate passenger travel data of a target route according to the spatial information of bus stops and the real data of the optimal candidate route. Compared with the traditional bus following method and the alighting stop inference method, the method can generate bus passenger travel data of a target route under the bus one-ticket system conveniently, quickly and at low cost based on the historical real data of the optimal candidate route. Only the spatial information of the stop is required. The method for generating data has high applicability and the obtained data has high accuracy, and can be applied to fields such as bus analysis.

[0072] Embodiment 2:

[0073] A system for generating single-line bus passenger travel data using the method for generating single-line bus passenger travel data in the above-mentioned embodiment 1 is provided, the system comprising:

[0074] The data acquisition module is used to randomly select G bus routes as candidate routes and obtain the historical riding data and station space information of all candidate routes;

[0075] A space vector module, used to determine the space vectors of all stations in the G candidate routes according to the station space information;

[0076] The spatial similarity total value calculation module is used to calculate the spatial similarity index between each station in the target line and all stations in any candidate line according to the spatial vector of the station, and obtain the total spatial similarity value between the candidate line and the target line. Then, each candidate line obtains a total spatial similarity value with the target line in this way;

[0077] A preferred candidate route screening module is used to sort the total spatial similarity values ​​of all candidate routes and the target route from large to small, and select the candidate routes corresponding to the first Q total spatial similarity values ​​as preferred candidate routes;

[0078] A data similarity total value calculation module is used to calculate the data similarity index between each station in the target line and all stations in any preferred candidate line, and obtain the data similarity total value between the preferred candidate line and the target line. Then, each preferred candidate line obtains a data similarity total value with the target line in this way.

[0079] The optimal candidate route screening module is used to calculate the similarity between each preferred candidate route and the target route according to the obtained total data similarity value, and select the preferred candidate route corresponding to the maximum similarity value as the optimal candidate route;

[0080] A generation module is used to use the optimal candidate route and the target route as learning samples and target samples, and to generate passenger travel data of the target sample from the route travel data of the learning sample through a cyclic generative adversarial network algorithm;

[0081] Among them, the data acquisition module, the space vector module, the space similarity total value calculation module, the preferred candidate line screening module, the data similarity total value calculation module, the optimal candidate line screening module, and the generation module communicate with each other wirelessly or wiredly.

[0082] Embodiment 3:

[0083] This embodiment is similar to Embodiment 2, except that the data acquisition module, space vector module, space similarity total value calculation module, preferred candidate line screening module, data similarity total value calculation module, optimal candidate line screening module, and generation module of the system in the above Embodiment 2 are integrated into a processor, and the simulation data finally output by the processor is displayed on a display screen so that the operator can intuitively observe the output results.

[0084] Obviously, the above embodiments of the present invention are only examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A method for generating single-line bus passenger travel data, characterized in that: The following steps are involved: S1. Randomly select G bus routes as candidate routes, and obtain the historical riding data and station space information of all candidate routes; S2. Determine the spatial vectors of all the sites in the G candidate routes according to the site spatial information; S3. According to the spatial vector of the station, the spatial similarity index of each station in the target line and all stations in any candidate line is calculated to obtain the total spatial similarity value between the candidate line and the target line; the calculation formula of the spatial similarity index is as follows: In the formula, s i is the i-th station in the target line, b j is the jth station in the candidate line, P(s i ,b j ) is site s i and site b k The spatial similarity index, l i 、h i For sites i The latitude and longitude and point of interest information, j 、h j For site b j The longitude and latitude and interest point information, α and β are the influence coefficients of the longitude and latitude and interest point information on the spatial similarity index, and the value range of the spatial similarity index P is between [-1,1]. The larger the value, the higher the spatial similarity of the two sites. S4. Each candidate route obtains a total spatial similarity value with the target route in the manner of step S3, sorts the total spatial similarity values ​​of all candidate routes with the target route from large to small, and selects the candidate routes corresponding to the first Q total spatial similarity values ​​as preferred candidate routes; S5. Calculate the data similarity index between each station in the target line and all stations in any preferred candidate line to obtain the total data similarity value between the preferred candidate line and the target line. The calculation formula of the data similarity index is as follows: In the formula, E(s i ,b j ) is site s i and site b j The data similarity index, R s , R b are the total number of passengers on the target route and the preferred candidate route, T u , T d are the stations of the target line respectively i The number of people getting on and off the bus, C u , C d Select the candidate line stations b respectively j The number of people getting on and off the bus, the data similarity index E ranges from [0,1]. The larger the value, the lower the data similarity of the two stations; S6. Each preferred candidate route obtains a total data similarity value with the target route in the manner of step S5. Based on the obtained total data similarity value, the similarity between each preferred candidate route and the target route is calculated, and the preferred candidate route corresponding to the maximum similarity value is selected as the optimal candidate route. The similarity calculation formula is as follows: In the formula, V(S,b) is the similarity between the preferred candidate route and the target route, E i ∈{E1,E2,…,E n }, E i is the site s in the target line i The maximum value of the data similarity index corresponding to the preferred candidate line, that is, is the total data similarity between the preferred candidate route and the target route. The value range of similarity V is between [0,1]. The larger the value, the more similar the preferred candidate route is to the target route. S7, taking the optimal candidate route and the target route as a learning sample and a target sample; S8. Generate the passenger travel data of the target sample using the route travel data of the learning sample through a cyclic generative adversarial network algorithm.

2. A method for generating single-line bus passenger travel data according to claim 1, characterized in that: In step S2, the site spatial information includes latitude and longitude information and interest point classification information. The process of determining the site spatial vector of the candidate route according to the site spatial information is as follows: Determine the amount of POI data under the longitude and latitude of each station and the tolerance radius, record the longitude and latitude information of each station as a 2-dimensional vector l, record the station's tolerable radius as r, and count the number of k types of POIs under the tolerable radius r of the station, recorded as a k-dimensional vector h. Then the spatial vector of each station is represented by a (2+k)-dimensional vector (l,h).

3. A method for generating single-line bus passenger travel data according to claim 2, characterized in that: In step S3, the specific process of obtaining the total spatial similarity value between the candidate route and the target route is as follows: Assume that the target line has n stations, that is, s i ∈{s1,s2,…,s n }, i = 1, 2, ..., n, the candidate line has m stations, that is, b j ∈{b1,b2,…,b m },j=1,2,…,m; First, calculate the spatial similarity index of the site s1 in the target line and the m sites in the candidate line, and obtain the m spatial similarity indexes corresponding to the site s1 and the m sites in the candidate line. Select the maximum value from them as the maximum value P1 of the spatial similarity index between the site s1 and the candidate line. Similarly, obtain the set of the maximum values ​​of the spatial similarity indexes between the n sites in the target line and the candidate line, recorded as {P1,P2,…,P n }, and finally P1, P2, ..., P n The sum is accumulated to obtain the total spatial similarity value between the candidate route and the target route.

4. The method for generating single-line bus passenger travel data according to claim 1, characterized in that: In step S5, the specific process of obtaining the total data similarity value between the preferred candidate route and the target route is as follows: First, the data similarity index of the site s1 in the target line and the m sites in the preferred candidate line is calculated to obtain the m data similarity indexes corresponding to the site s1 and the m sites in the preferred candidate line. The maximum value is selected as the maximum value E1 of the data similarity index corresponding to the site s1 and the preferred candidate line. Similarly, the set of the maximum values ​​of the data similarity indexes corresponding to the n sites in the target line and the preferred candidate line is obtained, which is recorded as {E1,E2,…,E n }, and finally E1, E2, …, E n The data are accumulated to obtain the total value of data similarity between the preferred candidate route and the target route.

5. The method for generating single-line bus passenger travel data according to claim 1, characterized in that: In step S7, the one-ticket riding dataset of the target route is used as the data source of the target samples, and the complete passenger riding dataset of the optimal candidate route is used as the data source of the learning samples.

6. A method for generating single-line bus passenger travel data according to claim 1, characterized in that: In step S8, a cyclic generative adversarial network model is used as a generator, the travel data matrix of the learning sample is input into the generator, simulation data is iteratively generated in the generator, and the generated simulation data is put into the judge for judgment. If the similarity of the simulation data is not lower than the set threshold, the generated simulation data is considered valid and output as the passenger travel data of the target sample.

7. A method for generating single-line bus passenger travel data according to claim 6, characterized in that: In the decision maker, the similarity between the simulated data and the real travel data of the learning sample is first calculated, and then the similarity is compared with the set threshold. If the similarity of the simulated data is not lower than the set threshold, the generated simulated data is considered valid and output; The set threshold refers to the similarity between the optimal candidate route obtained in step S6 and the target route.

Images