A method and system for generating route guidance information for urban rail transit

Abstract

This invention relates to a method and system for generating route guidance information for urban rail transit, specifically in the field of traffic flow guidance. The method includes: constructing a utility model for the display format of guidance information and a utility model for the content of guidance information based on information entropy, conditional entropy, and information gain; a route attribute includes multiple display formats, and each display format includes multiple display format levels; determining the perception coefficient of the route's guidance information based on the utility model for the display format and the utility model for the content of guidance information; constructing a passenger route selection model based on the perception coefficient, wherein the passenger route selection model is an improved Logit model; calibrating the model parameters of the passenger route selection model according to different types of passengers; and generating guidance information based on the passenger route selection models calibrated with different model parameters, with the goal of minimizing the uneven distribution of passenger flow. This invention alleviates the problem of uneven passenger flow distribution in the road network.

Description

Technical Field

[0001] This invention relates to the field of traffic flow guidance technology, and in particular to a method and system for generating urban rail transit route guidance information. Background Technology

[0002] With the continuous development of urban rail transit systems in major cities and the increasing passenger flow, higher demands are being placed on the networked operation of subways. Existing research shows that guidance information plays a significant role in guiding passengers' route selection behavior and is an important component of passenger transport organization methods. Currently, Chinese research focuses primarily on information content such as effective route search, with limited research on the patterns of passenger travel route selection behavior under the combined influence of guidance information content and its presentation. Furthermore, there is a lack of personalized guidance information generation methods adapted to the preferences of urban rail transit passengers. Therefore, there is an urgent need for a personalized generation method for urban rail transit route guidance information content and presentation to provide decision support for subway passenger flow guidance information generation. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for generating urban rail transit route guidance information, which alleviates the problem of uneven distribution of passenger flow on the road network.

[0004] To achieve the above objectives, the present invention provides the following solution:

[0005] A method for generating urban rail transit route guidance information includes:

[0006] Based on information entropy, conditional entropy, and information gain, we construct a utility model for the presentation form of induced information and a utility model for the content of induced information. The information entropy is the information entropy of the path selection decision event of urban rail transit passengers. The conditional entropy is the conditional entropy of the path selection decision event under the condition that a path attribute takes a certain presentation form level. A path attribute includes multiple presentation forms, and each presentation form includes multiple presentation form levels. The information gain is the information gain of the presentation form level of the path attribute on the path selection decision event.

[0007] Based on the utility model of the presentation format of the inducement information and the utility model of the content of the inducement information, the perceived coefficient of the inducement information of the path is determined;

[0008] A passenger route selection model is constructed based on the induced information perception coefficient, and the passenger route selection model is an improved Logit model;

[0009] The model parameters of the passenger route selection model are calibrated according to different types of passengers;

[0010] The passenger route selection model, calibrated based on different model parameters, generates guidance information with the goal of minimizing the uneven distribution of passenger flow.

[0011] Optionally, the information entropy is represented as:

[0012]

[0013] Where I represents the passenger's route selection decision event, i represents the route selection result, and i∈I R ,I R =(1,2,3,...), I R Indicates the set of alternative paths;

[0014] The conditional entropy is expressed as:

[0015] H(I|l)=-p(l)H(I|L f =l);

[0016] Among them, L f This indicates the horizontal display format of the path attribute f. This represents a horizontal set of display formats;

[0017] The information gain is expressed as: g(I,l)=H(I)-H(I|l).

[0018] Optionally, the route attributes include travel time attributes, congestion attributes, and transfer attributes;

[0019] The display formats include travel time display, congestion display, and transfer display. The travel time display is in text format, the congestion display includes text display, image display, and photo display, and the transfer display is in text format.

[0020] Optionally, the utility model of the inducement information display format is the sum of the normalized information gains of each path attribute corresponding to the set display format level.

[0021] The induced information content utility model is represented by the number of information particles and information accuracy, respectively.

[0022] The number of information particles is represented as: N R,a =1 / (ε'+Load) ave,a ),

[0023] in, Load u,v =q u,v / (N train ×λ train ), N R,a The number of information particles representing path a, ε' represents a non-zero positive number, Load ave,aI represents the average load factor of path a. R Represents the set of alternative paths, Load u,v q represents the load factor of the interval from node u to node v. u,v N represents the passenger flow through the interval from node u to node v. train This represents the number of trains passing through the interval from node u to node v, where node u and node v are points on path a, and λ is the number of trains passing through the interval from node u to node v. train This indicates the train's rated passenger capacity. This represents the total number of intervals along path a;

[0024] The accuracy of the information is expressed as: π R,a =1 / COV R ;

[0025] Among them, COV R COV represents the coefficient of variation of travel time. R =STD R / Mean R STD R Indicates OD to R(o) R ,d R Standard deviation of travel time within a day, Mean R Indicates OD to R(o) R ,d R The average daily travel time, o R Indicates the starting position, d R Indicates the destination location.

[0026] Optionally, determining the perceptual coefficient of the guiding information for the path based on the utility model of the guiding information display format and the utility model of the guiding information content specifically includes:

[0027] Based on the utility model of the presentation format of the inducement information and the utility model of the content of the inducement information, the utility of the inducement information of path a under the presentation format G is determined;

[0028] The perceptual coefficient of the induced information for path a is determined based on the utility of the induced information.

[0029] The utility of the inducement information is represented as: U R,a,G =N R,a ×π R,a ×H G ;

[0030] Among them, H G This represents the utility model of the presentation format of the induced information.

[0031] The inductive information perception coefficient is expressed as:

[0032] in

[0033] Where, k R,a,G Indicates OD to R(o) R ,d R The perceptual coefficient of G in the representation of path a, where e is a natural constant.

[0034] Optionally, the improved Logit model is represented as:

[0035]

[0036] Among them, P a This represents the probability of choosing path a. θ=[θ time ,θ load ,θ trans ], θ is a vector, θ time θ represents the travel time attribute coefficient. load θ represents the crowding attribute coefficient. trans This represents the transfer attribute coefficient. Let a be the vector representing the negative utility of path a. Let be the travel time for path a. Let γ represent the congestion level of path a, and γ represent the number of transfers for path a. To induce the information perception coefficient k R,a,G The function.

[0037] Optionally, the model parameters include travel time attribute coefficients, congestion attribute coefficients, and transfer attribute coefficients.

[0038] Alternatively, the objective function aimed at minimizing the uneven distribution of passenger flow can be expressed as:

[0039]

[0040] Where E represents the entropy of the road network load factor distribution in the designated area, g represents the discrete value of the load factor in the interval, and Ratio g This indicates the percentage of road network sections with a full load rate of g out of the total number of road network sections in the designated area;

[0041] The passenger route selection model, calibrated based on different model parameters, generates guidance information with the objective of minimizing the uneven distribution of passenger flow, specifically including:

[0042] For each type of passenger, a set of guidance information display schemes is determined by combining the display level corresponding to each path and the different path attribute display forms of each path. Each display scheme in the guidance information display scheme set includes the display level corresponding to the path attribute display form of each path.

[0043] Traverse the set of guidance information display schemes, calculate the path selection probability based on each guidance information display scheme according to the passenger path selection model, and determine the road network load factor distribution entropy of the set area based on the path selection probability.

[0044] The directional information display scheme corresponding to the minimum value among the road network load factor distribution entropies of multiple set areas obtained through traversal is used as the final directional information output.

[0045] This invention also discloses an urban rail transit route guidance information generation system, comprising:

[0046] The utility model construction module is used to construct utility models for induced information display formats and induced information content based on information entropy, conditional entropy, and information gain. The information entropy is the information entropy of the path selection decision event of urban rail transit passengers. The conditional entropy is the conditional entropy of the path selection decision event under the condition that a path attribute takes a certain display format level. A path attribute includes multiple display formats, and each display format includes multiple display format levels. The information gain is the information gain of the display format level of the path attribute on the path selection decision event.

[0047] The inducement information perception coefficient determination module is used to determine the inducement information perception coefficient of the path based on the inducement information display form utility model and the inducement information content utility model.

[0048] A passenger route selection model construction module is used to construct a passenger route selection model based on the induced information perception coefficient, wherein the passenger route selection model is an improved Logit model;

[0049] The model parameter calibration module is used to calibrate the model parameters of the passenger path selection model according to different types of passengers.

[0050] The guidance information generation module is used to generate guidance information based on the passenger route selection model calibrated with different model parameters, with the goal of minimizing the unevenness of passenger flow distribution.

[0051] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0052] This invention discloses a method and system for generating route guidance information for urban rail transit. Based on information entropy, conditional entropy, and information gain, it constructs a utility model for the presentation form of guidance information and a utility model for the content of guidance information, determines the perception coefficient of the route's guidance information, and constructs a passenger route selection model based on the perception coefficient. According to the passenger route selection model corresponding to different types of passengers, guidance information is generated with the goal of minimizing the uneven distribution of passenger flow. This achieves the modeling and characterization of passenger travel route selection behavior, alleviates the problem of uneven passenger flow distribution in the road network, and provides a reliable basis for generating passenger flow guidance information for urban rail transit. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This is a schematic diagram of the process for generating urban rail transit route guidance information according to the present invention;

[0055] Figure 2 This is a schematic diagram illustrating the display of passenger guidance information for different types during time period 1 in an embodiment of the present invention;

[0056] Figure 3 This is a schematic diagram illustrating the display of passenger guidance information for different types during time period 2 in an embodiment of the present invention;

[0057] Figure 4 This is a schematic diagram illustrating the display of passenger guidance information for different types during time period 3 in an embodiment of the present invention;

[0058] Figure 5 This is a schematic diagram of the objective function solution process according to an embodiment of the present invention;

[0059] Figure 6 This is a schematic diagram of the structure of an urban rail transit route guidance information generation system according to the present invention. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] The purpose of this invention is to provide a method and system for generating urban rail transit route guidance information, which alleviates the problem of uneven distribution of passenger flow on the road network.

[0062] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0063] Figure 1 This is a schematic diagram of the process for generating urban rail transit route guidance information according to the present invention, as shown below. Figure 1 As shown, a method for generating urban rail transit route guidance information includes:

[0064] Step 101: Construct a utility model for the presentation format of induced information and a utility model for the content of induced information based on information entropy, conditional entropy, and information gain; the information entropy is the information entropy of the path selection decision event of urban rail transit passengers, the conditional entropy is the conditional entropy of the path selection decision event under the condition that a path attribute takes a certain presentation format level, a path attribute includes multiple presentation formats, each presentation format includes multiple presentation format levels, and the information gain is the information gain of the presentation format level of the path attribute on the path selection decision event.

[0065] The route attributes include travel time attributes, congestion attributes, and transfer attributes.

[0066] The display formats include travel time display, congestion display, and transfer display. The travel time display is in text format, the congestion display includes text display, image display, and photo display, and the transfer display is in text format.

[0067] The display format levels are assigned as level 1, 2, 3, ... In actual guidance information, each path attribute adopts one of the display indicators or display styles, and multiple path attributes and their display format levels together constitute complete guidance information.

[0068] The information display formats for guiding information include display indicators and display styles for each path attribute; display indicators include range values ​​or average values; display styles include graphics, text, and real-scene photos.

[0069] The information provided in the guidance information includes route travel time, route congestion level, and number of transfers. The numerical values ​​of the indicators in the guidance information are averages from historical data of the same scenario and time period in the system database.

[0070] The information entropy of a path selection decision event is calculated using the information entropy calculation method. The information entropy is expressed as:

[0071]

[0072] Where I represents the passenger's route selection decision event, i represents the route selection result, and i∈I R ,I R =(1,2,3,...), I R This represents the set of alternative paths.

[0073] The conditional entropy of the path selection decision event under the condition that the path attribute f takes the display format level l is calculated based on the conditional entropy calculation method. The conditional entropy is expressed as:

[0074] H(I|l)=-p(l)H(I|L f =l);

[0075] Among them, L f This indicates the horizontal display format of the path attribute f. This represents a horizontal set of display formats;

[0076] The information gain of the formal level of a factor on the path selection decision event is calculated based on the information gain calculation method, which is the utility of the formal level l, expressed as: g(I,l)=H(I)-H(I|l).

[0077] The utility model for the induced information display format is the sum of the normalized information gains of each path attribute corresponding to the set display format G, taken as the level of the set display format, denoted as H. G ;

[0078] The induced information content utility model models information content from two dimensions: quantity and quality. The quantity dimension is represented by the number of information particles, and the quality dimension is represented by the accuracy of information.

[0079] The information particle number of path a is defined as the level of the number of passengers served by its current urban rail transit path, represented by the load factor.

[0080] The number of information particles is represented as: N R,a =1 / (ε'+Load) ave,a ),

[0081] in, N R,a The number of information particles representing path a, ε' represents a non-zero positive number, Load ave,a I represents the average load factor of path a. R Represents the set of alternative paths, Load u,v q represents the load factor of the interval from node u to node v. u,v N represents the passenger flow through the interval from node u to node v. trainThis represents the number of trains passing through the interval from node u to node v, where node u and node v are points on path a, and λ is the number of trains passing through the interval from node u to node v. train This indicates the train's rated passenger capacity. This represents the total number of intervals along path a.

[0082] Information accuracy is represented by travel time variability, and R(o) is calculated using the Kaparias method. R ,d R The coefficient of variation of travel time (COV) R .

[0083] COV R =STD R / Mean R ;

[0084] The accuracy of the information is expressed as: π R,a =1 / COV R ;

[0085] Among them, COV R STD represents the coefficient of variation of travel time. R Indicates OD to R(o) R ,d R Standard deviation of travel time within a day, Mean R Indicates OD to R(o) R ,d R The average daily travel time, o R Indicates the starting position, d R Indicates the destination location.

[0086] Step 102: Determine the perceptual coefficient of the path's guiding information based on the utility model of the guiding information display format and the utility model of the guiding information content.

[0087] Step 102 specifically includes:

[0088] Based on the utility model of the presentation format of the inducement information and the utility model of the content of the inducement information, the utility of the inducement information of path a under the presentation format G is determined.

[0089] The utility of the inducement information is represented as: U R,a,G =N R,a ×π R,a ×H G .

[0090] Among them, H G This represents the utility model of the presentation format of the induced information.

[0091] The induced information perception coefficient of path a is determined based on the induced information utility.

[0092] The inductive information perception coefficient is expressed as:

[0093] in

[0094] Where, k R,a,G Indicates OD to R(o) R ,d R The perceptual coefficient of G in the representation of path a, where e is a natural constant.

[0095] Step 103: Construct a passenger route selection model based on the induced information perception coefficient. The passenger route selection model is an improved Logit model.

[0096] Based on the induced information perception coefficient k R,a,G Adjusting the initial parameters of the polynomial Logit model:

[0097] θ=[θ time ,θ load ,θ trans ];

[0098]

[0099] θ is a vector, θ time θ represents the travel time attribute coefficient. load θ represents the crowding attribute coefficient. trans This represents the transfer attribute coefficient. It is called a specific alternative discrete parameter based on induced information perception, defined as the induced information perception coefficient k. R,a,G The function, These are intermediate parameters.

[0100] An improved Logit model is constructed based on the adjusted parameters:

[0101] The negative utility vector of travel path a, i.e., the cost feature vector, is:

[0102]

[0103] The impact of travel time, congestion, and number of transfers was measured. Let a be the vector representing the negative utility of path a. Let be the travel time for path a. Let γ represent the congestion level of path a, and γ represent the number of transfers for path a.

[0104] By the principle of utility maximization, the probability of choosing path a is:

[0105]

[0106] The improved Logit model is represented as follows:

[0107]

[0108] Among them, P a This represents the probability of choosing path a.

[0109] Step 104: Calibrate the model parameters of the passenger route selection model according to different types of passengers.

[0110] The model parameters include travel time attribute coefficients, congestion attribute coefficients, and transfer attribute coefficients.

[0111] Different types of passengers refer to the classification of passengers based on their personal attributes such as age, familiarity with the road network, and travel purpose. According to travel purpose, passengers can be divided into three categories: commuter / student passengers, daily rest passengers, and tourists. A survey was conducted on the route selection behavior of urban rail transit passengers under different forms of guidance information. The survey included: 1. A survey of passengers' personal socioeconomic attributes and travel habits; 2. A survey of passengers' route selection behavior under guidance information.

[0112] Based on the survey results, the maximum likelihood estimation method was used to calibrate the model parameters for different types of passenger distributions, and a personalized passenger route selection model under the influence of guidance information was constructed.

[0113] The survey on passenger travel route selection behavior under different forms of guidance information display refers to providing passengers with guidance information in different display formats, considering different levels of multiple route attributes. Specifically, as shown in Table 1, three levels of format were selected for each of the three route attributes included in the guidance information content: travel time, congestion level, and transfer situation; as shown in Table 2, an orthogonal experiment was designed with a total of nine combinations of guidance information display formats.

[0114] Table 1. Different display formats for each path attribute (factor)

[0115]

[0116] Table 2. Combination of Travel Time, Congestion Level, and Transfer Information Display Format

[0117]

[0118]

[0119] In this embodiment of the invention, the survey on passengers' personal socioeconomic attributes and travel habits refers to the survey of passengers' gender, age, weekly subway travel frequency, travel purpose, and travel route planning methods, in order to determine the comprehensiveness and rationality of the survey data.

[0120] In this embodiment of the invention, the survey of passenger route selection behavior under guidance information refers to: obtaining a set of alternative routes for travel origin-destination (OD) from historical data, and obtaining the values ​​of three route attributes—travel time, congestion level, and transfer status—for the corresponding time period from historical average flow factor (AFC) data. Under different information formats, travel route guidance information displaying the travel time, congestion level, and transfer status of each alternative route is provided. After selecting a route under different forms of guidance information, passengers click "Select this route," meaning each passenger participates in 9 experiments.

[0121] According to one embodiment of the present invention, a total of 438 passengers were surveyed. After removing unqualified records such as those with too short a completion time or contradictory options, 389 valid results were collected. Since each participant was required to make their own route selection under 9 types of APP guidance information schemes (as described above), the total dataset contains 3501 samples. The male-to-female ratio in the sample is basically balanced, with slightly more men than women; passengers aged 18 to 60 account for the majority; the majority of passengers travel 6 to 10 times per week; the vast majority of travelers plan their routes by querying information through a mobile APP; among the three historically common subway travel purposes—commuting to school, daily leisure, and travel—nearly half of the passengers were commuters to school, more than one-third were daily leisure passengers, and less than 20% were travelers. The data is similar to the samples collected in existing research surveys, indicating that the data collected in this survey is acceptable.

[0122] According to one embodiment of the present invention, the purpose of travel is used as the basis for classifying passenger types. The utility model of the inducement information display form based on information entropy, conditional entropy and information gain is constructed. The utility values ​​of different types of passengers are calculated by combining the information obtained from the survey data, as shown in Table 3.

[0123] Table 3 Utility Value Calculation Results for Different Types of Passengers

[0124]

[0125] According to an embodiment of the present invention, the parameter calibration results of the improved polynomial Logit model constructed by combining induced information perception coefficients are shown in Table 4.

[0126] Table 4. Parameter calibration results of the improved polynomial Logit model

[0127]

[0128] Table 5 Comparison between the improved MNL model and the basic MNL model

[0129]

[0130] Taking the route selection model for commuter and student passengers as an example, the comparison between the improved MNL model and the basic MNL model is shown in Table 5. As can be seen from Table 5, although the goodness of fit of both models is within the acceptable range, the improved MNL model (0.374) that considers the form of induced information is better than the unimproved MNL model (0.313).

[0131] Step 105: Based on the passenger route selection model calibrated with different model parameters, generate guidance information with the goal of minimizing the uneven distribution of passenger flow.

[0132] Aiming for the optimal balance of passenger flow distribution means using a road network status index that reflects the balance of passenger flow distribution, namely the road network load factor distribution entropy, to assess the road network status under passenger flow guidance.

[0133] The objective function, which aims to minimize the uneven distribution of passenger flow, is expressed as:

[0134]

[0135] Where E represents the entropy of the full load rate distribution of the road network in the set area. If the number of intervals included in the road network under study is D, then the value range of E is [0, lnD].

[0136] g represents the discrete value of the full load rate in the interval, with a precision of 0.1 and a maximum full load rate of 150%, i.e., g = (0, 0.1, 0.2, 0.3, ..., 1.5). The specific full load rate value is rounded to the discrete value according to the rounding rules.

[0137] Ratio g This indicates the percentage of road network sections with a full load rate of g out of the total number of road network sections in the designated area.

[0138] In a specific embodiment, step 105 specifically includes:

[0139] For each type of passenger, a set of guidance information display schemes is determined by combining the display level corresponding to each path and the different path attribute display forms of each path. Each display scheme in the guidance information display scheme set includes the display level corresponding to the path attribute display form of each path.

[0140] Each of the guidance information display schemes in the set is specifically a combination of travel time, congestion level, and transfers for each route.

[0141] Traverse the set of guidance information display schemes, calculate the path selection probability based on each guidance information display scheme according to the passenger path selection model, and determine the road network load factor distribution entropy of the set area based on the path selection probability.

[0142] The directional information display scheme corresponding to the minimum value among the road network load factor distribution entropies of multiple set areas obtained through traversal is used as the final directional information output.

[0143] In another specific embodiment, step 105 specifically includes:

[0144] Step 501: Information Scheme Initialization. Set the information scheme to G, set initial values ​​for the form level of each path utility influencing factor for each type of passenger, calculate the path travel cost and selection probability, and obtain the load factor distribution entropy by allocating traffic.

[0145] Step 502: Modify the level values ​​of each form of the guidance information for each type of passenger to obtain a new information scheme G', recalculate the route travel cost and selection probability under this scheme, and calculate the load factor distribution entropy after allocation.

[0146] Based on the aforementioned path selection model, passenger flow allocation under information scheme G is performed. Passenger flow allocation can be transformed into a Logit-based MSA model, satisfying the following constraints:

[0147]

[0148] In the formula: The planned trip OD is represented as R(o) R ,d R The number of passengers of type m who choose route a for their trip;

[0149] Represents the m-th type of passenger with respect to R(o) R ,d R ) OD demand;

[0150] The planned trip OD calculated by the route selection model is represented as R(o R ,d R The probability that the m-th type of passenger chooses route a for travel.

[0151] The passenger flow of each type of passenger on each route is superimposed to obtain the passenger flow of each section, the section load factor is calculated, and then the load factor distribution entropy is calculated.

[0152] Step 503: Compare the model target values ​​under G' and G. If the target value Z G >Z G' The optimal solution is updated to G', and the objective value is Z. G' Otherwise, retain the current optimal solution and target value unchanged.

[0153] Step 504: Determine whether the traversal of all information schemes and all passenger types is complete. If yes, output the optimal scheme G, the occupancy rate of each section, and the optimal target value Z.G Otherwise, return to step "Modify the level values ​​of each form of guidance information for each type of passenger to obtain a new information scheme G', recalculate the route travel cost and selection probability under this scheme, and calculate the load factor distribution entropy after allocation".

[0154] Step 505: Output the optimal guidance information scheme and optimal target Z for various types of passengers, that is, generate the guidance information content and display format that minimizes the target value.

[0155] like Figure 5 The flowchart shown illustrates the solution process for the objective function, which includes the following steps:

[0156] Step 1: Initialization. Assign initial values ​​to each variable, and input the urban rail transit network topology data, normal network clearing data, and AFC data under the same historical station closure events during the research period.

[0157] Step 2: Clean the AFC passenger flow data, extract the OD passenger flow matrix of the network under study, and denote the OD to R(o R ,d R The demand is q. R Assuming that demand does not shift to other modes of transportation during station closures, i.e., demand remains unchanged;

[0158] Step 3: Statistics and recording of effective paths.

[0159] Step 4: Let the current assigned passenger be of type m, and set the initial value m = 1.

[0160] Step 5: Based on the proportion of passengers' travel purposes obtained from the survey, calculate the OD demand of the m-th type of passengers; Indicates OD to R(o) R ,d R The proportion of passengers in the m-th class among ) is then OD to R(o R ,d R The travel demand of passenger type m is

[0161] Step 6: Initialize the information format scheme. Set the information scheme to... G For each type of passenger, the formal level of each path utility influencing factor is set to an initial value (1, 1, 1), the path travel cost and selection probability are calculated, and the load factor distribution entropy is obtained by allocating the flow.

[0162] Step 7: Modify the level values ​​of each form of the guidance information for each type of passenger to obtain a new information scheme G', recalculate the route travel cost and selection probability under this scheme, and calculate the load factor distribution entropy after allocation.

[0163] Step 8: Compare the model target values ​​under G' and G. If the target value Z G >Z G' The optimal solution is updated to G', and the objective value is Z. G' Otherwise, retain the current optimal solution and target value unchanged.

[0164] Step 9: Determine if the traversal of all information schemes and all passenger types is complete. If yes, proceed to Step 10 and output the optimal scheme G, the load factor of each section, and the optimal target value Z. G Otherwise, return to Step 7.

[0165] Step 10: Output the optimal guidance information scheme and the optimal target Z for each type of passenger.

[0166] In this embodiment, the scenario of uneven passenger flow distribution caused by a station closure event is selected. Passenger flow guidance information is generated using three time periods as examples, and the degree of balance in road network passenger flow distribution during this time period is compared between a normal day and a day with or without generated guidance information, as shown in Table 6. The levels 1, 2, and 3 in the information format scheme in the table have the same meaning as in Table 1. Figure 2-4 This illustration shows the content and interface of the guidance information generated for commuting and school-going passengers in the embodiment. Figure 2 , Figure 3 and Figure 4 In Table 6, (a) represents commuter passengers, (b) represents leisure passengers, and (c) represents tourists. As shown in Table 6, under the influence of the guidance information, the difference in network load factor distribution entropy between closed days and normal days decreased by 42%–70%, and the generated guidance information effectively reduced the uneven distribution of passenger flow on the network.

[0167] Table 6. Degree of Balance in Passenger Flow Distribution on the Road Network During Different Time Periods When Guidance Information is Generated

[0168]

[0169] This invention discloses a method for generating urban rail transit route guidance information, specifically a personalized method for generating urban rail transit route guidance information content and display format. The method includes: constructing a utility model for the display format of guidance information based on information entropy, conditional entropy, and information gain; constructing an information content utility model; calculating the perception coefficient of guidance information under the combined utility of content and format; constructing an urban rail transit passenger route selection model under the influence of guidance information based on the perception coefficient, and calibrating model parameters for different types of passengers; and establishing a personalized guidance information content and format generation model with the goal of minimizing the uneven distribution of passenger flow. This invention achieves quantitative analysis of the impact of guidance information display format on the travel route selection behavior of urban rail transit passengers; it integrates route attribute factors such as passenger travel time, congestion, and number of transfers with guidance information to model and characterize passenger travel route selection behavior; the generated guidance information effectively alleviates the problem of uneven distribution of passenger flow in the road network and provides a reliable basis for the generation of urban rail transit passenger flow guidance information.

[0170] The beneficial effects of this invention are as follows:

[0171] 1. This invention proposes a method for measuring the form perception of induced information based on information entropy, conditional entropy, and information gain, calculating the induced information perception coefficient by comprehensively considering information content and form utility. This invention quantifies the impact of information form level on the route selection behavior of urban rail transit passengers, providing a theoretical basis for comprehensively characterizing the passenger route selection patterns under induced information.

[0172] 2. This invention proposes an improved multinomial Logit passenger travel route selection behavior modeling method that considers the influence of directional information content and display format. Compared with the unimproved multinomial Logit model method that does not consider the influence of information format, the improved model has better fit and higher reliability. This invention provides effective methodological support for understanding the travel route selection patterns of different types of subway passengers under the influence of directional information.

[0173] 3. This invention proposes a method for personalized generation of urban rail transit route guidance information content and display format. The method considers the impact of information content and format on the route selection behavior of different types of passengers, and the generated guidance information effectively reduces the level of road network imbalance, providing a reliable theoretical basis for formulating passenger flow guidance information.

[0174] Figure 6 This is a schematic diagram of the structure of an urban rail transit route guidance information generation system according to the present invention, such as... Figure 6 As shown, an urban rail transit route guidance information generation system includes:

[0175] The utility model construction module 201 is used to construct a utility model of the induced information display form and a utility model of the induced information content based on information entropy, conditional entropy and information gain; the information entropy is the information entropy of the path selection decision event of urban rail transit passengers, the conditional entropy is the conditional entropy of the path selection decision event under the condition that a path attribute takes a display form level, a path attribute includes multiple display forms, each display form includes multiple display form levels, and the information gain is the information gain of the display form level of the path attribute on the path selection decision event;

[0176] The inducement information perception coefficient determination module 202 is used to determine the inducement information perception coefficient of the path based on the inducement information display form utility model and the inducement information content utility model;

[0177] The passenger route selection model construction module 203 is used to construct a passenger route selection model based on the induced information perception coefficient, wherein the passenger route selection model is an improved Logit model;

[0178] The model parameter calibration module 204 is used to calibrate the model parameters of the passenger path selection model according to different types of passengers.

[0179] The guidance information generation module 205 is used to generate guidance information based on the passenger route selection model calibrated with different model parameters, with the goal of minimizing the unevenness of passenger flow distribution.

[0180] A system for generating route guidance information for urban rail transit includes: a route selection behavior prediction module, a road network status assessment module, an information generation module, and a data management module.

[0181] The route selection behavior prediction module is used to calculate the route selection probability by taking route attribute index data and guidance information content and form scheme under the same historical scenario and time period as input, and applying the aforementioned guidance information-based route selection model that considers passenger personalization.

[0182] The road network status assessment module is used to allocate passenger flow and calculate road network status indicators based on the results of the route selection module as input, and is used to evaluate the guidance effect brought about by the aforementioned input guidance information scheme.

[0183] The information generation module is used to generate a combination of directional information content and format that optimizes the road network state and satisfies the perceptual preferences of various passengers, thereby providing personalized route guidance information for urban rail transit passengers.

[0184] The data management module is used to provide basic data support for the system, including basic road network data, historical route attribute data, feasible route set data, passenger account information, and information templates.

[0185] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0186] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for generating urban rail transit route guidance information, characterized in that, include: Based on information entropy, conditional entropy, and information gain, construct utility models for the presentation of induced information and the content of induced information; The information entropy is the information entropy of the path selection decision event of urban rail transit passengers. The conditional entropy is the conditional entropy of the path selection decision event under the condition that a path attribute takes a certain display form level. A path attribute includes multiple display forms, and each display form includes multiple display form levels. The information gain is the information gain of the display form level of the path attribute on the path selection decision event. Based on the utility model of the presentation format of the inducement information and the utility model of the content of the inducement information, the perceived coefficient of the inducement information of the path is determined; A passenger route selection model is constructed based on the induced information perception coefficient, and the passenger route selection model is an improved Logit model; The model parameters of the passenger route selection model are calibrated according to different types of passengers; Based on the passenger route selection model calibrated with different model parameters, guidance information is generated with the goal of minimizing the uneven distribution of passenger flow. The utility model for the induced information display format is the sum of the normalized information gains of each path attribute corresponding to the set display format level. The induced information content utility model is represented by the number of information particles and information accuracy, respectively. The number of information particles is represented as: ; in, , , Representing a path a The number of information particles, Represents a non-zero positive number. Representing a path a Average load factor Represents the set of alternative paths. Indicates from node u To the node v The load factor of the interval Indicates from node u To the node v Passenger flow through the section Indicates from node u To the node v The number of trains passing through the section, nodes u To the node v For path a The point on, This indicates the train's rated passenger capacity. Representing a path a The total number of intervals above; The accuracy of the information is expressed as follows: ; in, Indicates the coefficient of variation of travel time. , Indicates OD pair Standard deviation of travel time within a day Indicates OD pair The average travel time within a day Indicates the starting position. Indicates the destination location; Based on the utility model of the presentation format of the inducement information and the utility model of the content of the inducement information, the perceived coefficient of the inducement information of the path is determined, specifically including: The path is determined based on the utility model of the presentation format of the inducement information and the utility model of the content of the inducement information. a In terms of presentation format G The persuasive effect of information provided; Determine the path based on the effectiveness of the inducing information. a The coefficient of perception of induced information; The utility of the inducing information is represented as follows: ; in, This represents the utility model of the presentation format of the induced information; The inductive information perception coefficient is expressed as: ； in, Indicates OD pair path The perceptual coefficient of the display format G, It is a natural constant; The improved Logit model is represented as follows: ； in, Representing a path The probability of choosing, , , , , For vectors, This represents the travel time attribute coefficient. This represents the crowding level attribute coefficient. This represents the transfer attribute coefficient. Representing a path The negative utility vector, For path Travel time, For path The level of congestion, For path Number of transfers, To induce information perception coefficient The function.

2. The method for generating urban rail transit route guidance information according to claim 1, characterized in that, The information entropy is represented as: ； in, I This represents a passenger's route selection decision event. i This indicates the path selection result. , Indicates the set of alternative paths; The conditional entropy is expressed as: ； in, Represents path attributes f The level of the presentation format , This represents a horizontal set of display formats; The information gain is expressed as: .

3. The method for generating urban rail transit route guidance information according to claim 1, characterized in that, The route attributes include travel time attributes, congestion attributes, and transfer attributes; The display formats include travel time display, congestion display, and transfer display. The travel time display is in text format, the congestion display includes text display, image display, and photo display, and the transfer display is in text format.

4. The method for generating urban rail transit route guidance information according to claim 1, characterized in that, The model parameters include travel time attribute coefficients, congestion attribute coefficients, and transfer attribute coefficients.

5. The method for generating urban rail transit route guidance information according to claim 4, characterized in that, The objective function, which aims to minimize the uneven distribution of passenger flow, is expressed as: ； in, E This represents the entropy of the road network load factor distribution in a given area. g This represents the discrete value of the load factor for a given interval. This indicates the full load rate of the total number of road network sections in the designated area. The percentage of the interval; The passenger route selection model, calibrated based on different model parameters, generates guidance information with the objective of minimizing the uneven distribution of passenger flow, specifically including: For each type of passenger, a set of guidance information display schemes is determined by combining the display level corresponding to each path and the different path attribute display forms of each path. Each display scheme in the guidance information display scheme set includes the display level corresponding to the path attribute display form of each path. Traverse the set of guidance information display schemes, calculate the path selection probability based on each guidance information display scheme according to the passenger path selection model, and determine the road network load factor distribution entropy of the set area based on the path selection probability. The directional information display scheme corresponding to the minimum value among the road network load factor distribution entropies of multiple set areas obtained through traversal is used as the final directional information output.

6. A system for generating route guidance information for urban rail transit, characterized in that, include: The utility model construction module is used to construct utility models of the form of induced information display and utility models of induced information content based on information entropy, conditional entropy and information gain. The information entropy is the information entropy of the path selection decision event of urban rail transit passengers. The conditional entropy is the conditional entropy of the path selection decision event under the condition that a path attribute takes a certain display form level. A path attribute includes multiple display forms, and each display form includes multiple display form levels. The information gain is the information gain of the display form level of the path attribute on the path selection decision event. The inducement information perception coefficient determination module is used to determine the inducement information perception coefficient of the path based on the inducement information display form utility model and the inducement information content utility model. A passenger route selection model construction module is used to construct a passenger route selection model based on the induced information perception coefficient, wherein the passenger route selection model is an improved Logit model; The model parameter calibration module is used to calibrate the model parameters of the passenger path selection model according to different types of passengers. The guidance information generation module is used to generate guidance information based on the passenger route selection model calibrated with different model parameters, with the goal of minimizing the unevenness of passenger flow distribution. The utility model for the induced information display format is the sum of the normalized information gains of each path attribute corresponding to the set display format level. The induced information content utility model is represented by the number of information particles and information accuracy, respectively. The number of information particles is represented as: ; in, , , Representing a path a The number of information particles, Represents a non-zero positive number. Representing a path a Average load factor Represents the set of alternative paths. Indicates from node u To the node v The load factor of the interval Indicates from node u To the node v Passenger flow through the section Indicates from node u To the node v The number of trains passing through the section, nodes u To the node v For path a The point on, This indicates the train's rated passenger capacity. Representing a path a The total number of intervals above; The accuracy of the information is expressed as follows: ; in, Indicates the coefficient of variation of travel time. , Indicates OD pair Standard deviation of travel time within a day Indicates OD pair The average travel time within a day Indicates the starting position. Indicates the destination location; Based on the utility model of the presentation format of the inducement information and the utility model of the content of the inducement information, the perceived coefficient of the inducement information of the path is determined, specifically including: The path is determined based on the utility model of the presentation format of the inducement information and the utility model of the content of the inducement information. a In terms of presentation format G The persuasive effect of information provided; Determine the path based on the effectiveness of the inducing information. a The coefficient of perception of induced information; The utility of the inducing information is represented as follows: ; in, This represents the utility model of the presentation format of the induced information; The inductive information perception coefficient is expressed as: ； in, Indicates OD pair path The perceptual coefficient of the display format G, It is a natural constant; The improved Logit model is represented as follows: ； in, Representing a path The probability of choosing, , , , , For vectors, This represents the travel time attribute coefficient. This represents the crowding level attribute coefficient. This represents the transfer attribute coefficient. Representing a path The negative utility vector, For path Travel time, For path The level of congestion, For path Number of transfers, To induce information perception coefficient The function.

Images