A personalized regional travel scheme planning method considering travel time reliability

Abstract

The application provides a personalized regional travel scheme planning method considering travel time reliability. The method comprises the following steps: initializing regional traffic supply information, inputting regional traffic travel information; calculating the travel time budget of each road section for the traveler i; generating a regional travel initial selection limb set; calculating the travel tolerance time threshold; determining the regional travel effective selection limb set; calculating the selection probability of the regional travel effective selection limb; and generating a personalized regional travel planning scheme in seven steps. The application is based on a combined travel mode and path joint selection potential classification model under time uncertainty, realizes comprehensive consideration of various factors and in-depth consideration of regional travel heterogeneity, introduces the influence of travel time uncertainty on the selection set generation and traveler selection preference, provides a "urban-intra-urban-intercity" whole process regional travel planning scheme for different travelers, and meets the personalized regional travel needs of passengers.

Description

Technical Field

[0001] This invention relates to the field of regional technology, and in particular to a personalized regional travel planning method that takes into account the reliability of travel time. Background Technology

[0002] As urban agglomerations and metropolitan areas expand their transportation networks at various levels within cities and blocks, and between different cities and blocks, the connections between these networks are becoming increasingly complex. This has led to more pronounced multi-modal travel characteristics and greater uncertainty in travel times. Furthermore, the diversification of regional travel activities and the heterogeneity of residents' travel behaviors necessitate a planning method for personalized and reliable travel solutions at the regional scale that takes into account the uncertainty of travel times. This will facilitate efficient, convenient, and reliable personalized regional travel combinations.

[0003] Regional combined travel encompasses the entire journey from origin to destination, including "intra-city-intercity-intra-city" travel. Taking travel between Beijing and Tianjin as an example, it generally consists of three parts: initial intra-city travel, intercity travel, and final intra-city travel, involving different levels and multiple modes of transportation, resulting in various combinations of regional travel modes. Compared to intra-city or intercity travel, regional combined travel offers a larger range of travel options, longer travel distances, and more uncertain travel times. Furthermore, due to differences in socioeconomic attributes, risk attitudes, and travel purposes, travelers often prefer different travel plans. These phenomena pose a significant challenge to generating personalized and reliable travel plans at the regional scale.

[0004] Currently, the shortcomings of existing regional travel planning methods include: neglecting the joint planning of urban transportation modes when focusing on intercity trains or air travel; providing planning solutions with the goal of shortest time or lowest cost, ignoring travelers' comprehensive consideration of various factors; ignoring the impact of travel time uncertainty on travel planning and travelers' travel preferences; and ignoring the differences in travelers' travel preferences. Summary of the Invention

[0005] This invention provides a personalized regional travel planning method that considers the reliability of travel time, so as to provide different travelers with a full-process regional travel planning scheme of "intra-city-intercity" and meet the personalized regional travel needs of passengers.

[0006] To achieve the above objectives, the present invention adopts the following technical solution.

[0007] A personalized regional travel planning method that considers the reliability of travel time includes:

[0008] Initialize regional traffic supply information and regional traffic travel information, obtain the personal attributes and travel mode attributes of travelers, and calculate the travel time budget for each road segment for traveler i.

[0009] Based on the origin and destination O of each route segment for traveler i i and D i Given a set N of combined regional travel modes, calculate the initial path selection set K for each combined regional travel mode using the K-shortest path algorithm. n ;

[0010] Obtain the average travel time t of traveler i choosing different regional combinations of travel modes n. in Determine the shortest travel time for traveler i when choosing different regional combinations of travel modes n. And calculate its travel time tolerance threshold T in ; Obtain the average travel time t of traveler i choosing travel route k under the condition that the traveler i selects regional combination travel mode n. ink Given that traveler i selects a combination of regional travel modes n, determine the shortest travel time for choosing travel route k. And calculate its travel time tolerance threshold T ink ;

[0011] T in With the t in Compare and filter effective regional combinations of travel modes; [the T] ink With the Compare and filter effective travel options to generate a set G of effective regional combinations of travel modes and effective options for effective travel options.

[0012] Calculate the probability that traveler i belongs to potential category s, calculate the probability that traveler i chooses the n′ combination of travel modes and the effective travel path k′ when belonging to potential category s, and then calculate the probability that traveler i chooses each effective option g:

[0013] Based on the probability of traveler i choosing each valid option g, sort and output the joint set of regional combination travel modes and valid selected routes, and output the personalized travel plan recommendation results for traveler i, including combination travel mode, travel route, estimated time and cost, estimated departure time and estimated arrival probability.

[0014] Preferably, the initialization of regional traffic supply information and regional traffic travel information, and the acquisition of travelers' personal attributes and travel mode attributes, include:

[0015] Obtain multi-level rail transit operation plans and bus network operation plans for the loaded area, including the travel time distribution of urban roads and highways at all times of the day, the rail transit station entry and exit time distribution, the transfer time distribution within the same rail transit system and between different rail transit systems, as well as the travel cost (Cost) and the mean t and variance t of the travel time for each mode of transportation and each travel route. SD information;

[0016] Obtain the traveler's personal attributes and travel mode attributes. Personal attributes include age, occupation, education, and monthly income. Travel mode attributes include the regional origin O. i Destination D in the region i Purpose of travel P i Information on expected arrival probability ε and estimated arrival time.

[0017] Preferably, the calculation of the travel time budget for each segment of traveler i includes:

[0018] Assuming that the travel time data of each traveler are independent, we fit the travel time data of each traveler to a normal distribution, and obtain the mean t and standard deviation t of the travel time for each road segment under different modes of transportation. SD ;

[0019] Based on the transformation of its utility function, the formula for calculating travel time budget B for traveler i can be derived as follows:

[0020]

[0021] In the formula, β t and β SD These are the coefficients of the mean travel time t and the standard deviation travel time t in the utility function of the mean-standard deviation NL model, respectively. SD The coefficient;

[0022] Under the desired conditions, the mathematical relationship between the expected arrival time probability ε and the travel time budget B is as follows:

[0023]

[0024] t α Indicates the traveler's actual travel time;

[0025] The formula can be transformed to obtain:

[0026]

[0027] RR represents the travel time risk coefficient, which indicates a traveler's attitude towards the trade-off between the mean travel time and the uncertainty of travel time.

[0028] Based on the above formula, the RR (Return Rate) β of traveler i is calculated according to the expected arrival time probability ε.SD / β t Combining the mean t and standard deviation t of travel time for each road segment under different modes of transportation SD Calculate the travel time budget B for each segment of traveler i.

[0029] Preferably, the origin and destination O of each road segment based on traveler i are... i and D i Given a set N of combined regional travel modes, calculate the initial path selection set K for each combined regional travel mode using the K-shortest path algorithm. n ,include:

[0030] Based on the origin and destination O of each route segment for traveler i i and D i Determine the set N of regional travel modes;

[0031] For any regional combination of travel modes n in the set N of combined travel modes, the K-shortest path algorithm is used to determine the initial set of selected paths K under the combined travel modes in that region, with travel time budget as the indicator. n ;

[0032] Combine the regional travel mode set N with the initial selected path set K n Perform joint generation of the initial selection set for regional travel.

[0033] Preferably, the step of obtaining the average travel time t of traveler i choosing different regional combinations of travel modes n is... in Determine the shortest travel time for traveler i when choosing different regional combinations of travel modes n. And calculate its travel time tolerance threshold T in ; Obtain the average travel time t of traveler i choosing travel route k under the condition that the traveler i selects regional combination travel mode n. ink Given that traveler i selects a combination of regional travel modes n, determine the shortest travel time for choosing travel route k. And calculate its travel time tolerance threshold T ink ,include:

[0034] Based on the travel origin O of traveler i i and endpoint D i Determine the average travel time t of traveler i when choosing different regional combinations of travel modes n. in Extract the shortest travel time for traveler i when choosing different combinations of travel modes in different regions.

[0035] Construct a traveler tolerance time threshold function:

[0036] Given that traveler i selects a combination of regional travel modes n, the average travel time t for choosing travel route k is determined. ink Given that traveler i selects a combination of regional travel modes n, determine the shortest travel time for the chosen travel route k.

[0037] Construct a traveler tolerance time threshold function:

[0038] Preferably, the T in With the t in Compare and filter effective regional combinations of travel modes; [the T] ink With the t ink By comparing and filtering effective travel options, a set G of effective choices for effective regional combinations of travel modes and effective routes is generated, including:

[0039] The travel time tolerance threshold T for regional combination travel modes in The average travel time t of traveler i choosing different regional combinations of travel modes n in Compare and select effective regional combinations of travel modes;

[0040] The travel time tolerance threshold T for choosing travel route k under the condition that traveler i selects a combination of regional travel modes n. ink Given that traveler i selects a combination of regional travel modes n, the average travel time t for choosing travel route k. ink Compare and filter the effective travel options;

[0041] Generate a joint choice set G = {G1; G2; ... G} of effective regional travel mode combinations and effective route selections. n′}, where G1={G 11 G 12 …G 1k′} represents the set of all valid choice paths k′ for the effective combination of regional travel modes n′.

[0042] Preferably, the calculation of the probability that traveler i belongs to potential category s, the calculation of the probability that traveler i chooses the n′ combination of travel modes and the effective travel path k′ when belonging to potential category s, and the calculation of the probability that traveler i chooses each effective option g, includes:

[0043] Construct a latent classification model for travelers, and use this model to divide travelers into different latent categories. The probability that traveler i belongs to latent category s is:

[0044]

[0045] Vis =ASC is +∑η is ·X is

[0046] In the formula, V is For travelers i, this is a fixed term in the utility function when i belongs to the potential category s; For all potential class utilities and; ASC is X is a constant term when traveler i belongs to the potential category s; is Factors influencing whether traveler i belongs to potential category s include age, vehicle ownership, gender, income, and occupation. is For the corresponding parameters;

[0047] The utility function of the mean-standard deviation NL model is:

[0048]

[0049] In the formula, V ig ASC is a fixed term in the utility function when traveler i selects an effective choice limb g; ig For traveler i, the constant term is the effective choice limb g; t is the mean travel time for traveler i when choosing the effective choice limb g; t SD For traveler i, select the standard deviation of travel time under the effective choice limb; X ig Factors influencing traveler i's choice of effective limb g; η ig For the corresponding parameters;

[0050] The probability of traveler i choosing the n′ combination of travel modes and the effective travel path k′ is calculated based on the mean-standard deviation NL model.

[0051]

[0052] In the formula, X isn′k′ Factors influencing traveler i's choice of the n′ regional combination travel mode and selection of the effective route k′, belonging to potential category s, include travel purpose, travel cost, travel distance, number of transfers, and familiarity with the road network. X is an influencing factor in selecting an effective route k′ given that traveler i belongs to potential category s and chooses regional combination travel mode n′. isn′ For travelers i belonging to potential category s, the factors influencing their choice of regional combination travel mode n′ are: η isn′k′ , η isn′ For the corresponding parameters;

[0053] Calculate the probability that traveler i chooses each valid option g:

[0054]

[0055] In the formula, LOGSUM represents the combined utility term of traveler i choosing the n′th effective regional combination travel mode and the lower-level travel route selection limb, and μ is the scaling coefficient connecting the upper and lower layers of the NL model.

[0056] Preferably, the step of sorting and outputting a joint set of regional combination travel modes and effective selected routes based on the probability of traveler i selecting each effective option g, and outputting a personalized travel plan recommendation result for traveler i, includes combination travel modes, travel routes, estimated travel time and cost, estimated departure time and estimated arrival probability, including:

[0057] The output is a joint set of regional travel modes and effective routes, sorted according to the selection probabilities of each effective option. Based on the time budget of each route under each regional travel mode, and combined with the β value obtained from the latent classification NL model indicating whether the traveler belongs to different latent categories... t The parameter values ​​are used to calculate β corresponding to the effective choice made by traveler i. SD β t Values;

[0058] The travel time risk coefficients of travelers belonging to each potential category and choosing a certain effective option are weighted and summed according to the probability of the traveler belonging to each potential category to obtain the travel time risk coefficient of the traveler choosing a certain effective option; the on-time arrival probability of the traveler choosing a certain effective option is calculated based on the travel time risk coefficient of the traveler choosing a certain effective option.

[0059] The system filters travel options whose on-time arrival probability ξ satisfies the expected on-time arrival probability ε and whose effective selection probability is greater than the threshold R. The options are sorted according to their selection probability, and the first set number of travel options are output as the personalized travel option recommendation results for traveler i. These options include the combination of travel modes, travel route, estimated time and cost, estimated departure time and estimated arrival probability.

[0060] As can be seen from the technical solutions provided by the embodiments of the present invention above, the present invention proposes a personalized regional travel plan planning method that considers the reliability of travel time. Based on the potential classification model of combined travel modes and routes under time uncertainty, it realizes the comprehensive consideration of multiple factors and the in-depth consideration of regional travel heterogeneity. It also introduces the impact of travel time uncertainty on the generation of choice sets and traveler's choice preferences, providing different travelers with a full-process regional travel planning scheme of "intra-city-intercity" to meet the personalized regional travel needs of passengers.

[0061] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and will become apparent from the description or may be learned by practice of the invention. Attached Figure Description

[0062] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of 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.

[0063] Figure 1 A flowchart of a personalized regional travel plan planning method considering travel time reliability is provided in an embodiment of the present invention;

[0064] Figure 2 A schematic diagram of the physical topology of a regional multi-level transportation super network provided in an embodiment of the present invention;

[0065] Figure 3 A flowchart illustrating the design of a K-shortest path algorithm provided in this embodiment of the invention;

[0066] Figure 4 This is a recommended solution corresponding to an effective path provided in an embodiment of the present invention;

[0067] Figure 5 This is an effective path corresponding to a recommended scheme two provided in an embodiment of the present invention;

[0068] Figure 6 This is a recommended solution three corresponding to an effective path provided in an embodiment of the present invention. Detailed Implementation

[0069] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0070] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or couplings. The term “and / or” as used herein includes any and all combinations of one or more of the associated listed items.

[0071] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0072] To facilitate understanding of the embodiments of the present invention, the following will provide further explanation and description with reference to the accompanying drawings and several specific embodiments. These embodiments do not constitute a limitation on the embodiments of the present invention.

[0073] This invention utilizes a latent classification NL model to explore the preferences and heterogeneity of combined regional travel modes and routes under uncertain travel times. By comprehensively considering multiple influencing factors, it proposes a personalized regional travel plan method that takes into account the reliability of travel times. This invention latently classifies travelers' personal attributes, analyzes the travel characteristics of potential categories, and constructs a joint selection behavior model of shortest path and combined regional travel mode choices, providing better personalized information services for regional travel.

[0074] The processing flow of a personalized regional travel plan planning method considering travel time reliability provided in this embodiment of the invention is as follows: Figure 1 As shown, the processing steps include the following:

[0075] Step S1: Initialize regional traffic supply information and regional traffic travel information, and obtain the personal attributes and travel mode attributes of travelers.

[0076] The physical topology and traffic status of a multi-level traffic supernetwork in a loading area provided by this invention are as follows: Figure 2As shown, the multi-level rail transit operation plan and bus network operation plan of the loaded area are obtained, including the travel time distribution of urban roads and highways at all times of the day, the rail transit station entry and exit time distribution, and the transfer time distribution within the same rail transit system and between different rail transit systems, as well as the travel cost (Cost) and the mean t and variance t of the travel time for each mode of transportation and each travel route. SD Basic information such as the start and end points of the rows. Use Python to crawl and extract these information. i and D i The travel time data for all time periods across all travel routes involved in regional travel are aggregated and fitted with a normal distribution to obtain the mean t and variance t of the travel time for each travel route. SD .

[0077] Obtain the traveler's personal attributes and travel mode attributes. Personal attributes include age, occupation, education, and monthly income, while travel mode attributes include the regional origin O. i Destination D in the region i Purpose of travel P i Information such as expected probability of arrival ε and estimated arrival time. When setting the parameter type for input data, personal attributes are set as dummy variables. Dummy variables are numerical variables, typically taking values ​​of 0 or 1.

[0078] Table 1 is a schematic table of personal attributes and travel mode attributes of a traveler provided in an embodiment of the present invention.

[0079]

[0080]

[0081] Step S2: Calculate the travel time budget for each route segment for traveler i.

[0082] Travel time data fitting: The travel times for each segment of the route to the major railway hubs within the city are summarized across all time periods for each mode of transportation. Assuming that the travel time data for each traveler are independent, the travel time data are fitted to a normal distribution, yielding the mean t and standard deviation t0 for each segment under different modes of transportation. SD Mean and variance t, t0 of travel time distribution data under different transportation modes and different route choices SD different.

[0083] Based on the transformation of its utility function, the time budget B can be derived:

[0084] The formula for calculating travel time budget B is as follows:

[0085]

[0086] βt and β SD What does mean?

[0087] In the formula, β t and β SD These are the coefficients of the mean travel time t and the standard deviation travel time t in the utility function of the mean-standard deviation NL model, respectively. SD The coefficient;

[0088] Under the expected conditions, the expected on-time arrival probability ε is the same as the on-time arrival probability ξ under the expected conditions and follows a standard normal distribution. The mathematical relationship between the expected arrival time probability ε and the travel time budget B is as follows:

[0089]

[0090] t α This indicates the traveler's actual travel time.

[0091] The formula can be transformed to obtain:

[0092]

[0093] RR represents the travel time risk coefficient, which reflects a traveler's attitude towards the trade-off between the mean travel time and the uncertainty of travel time.

[0094] Based on the above formula, given the expected arrival time probability ε, calculate the traveler's RR, i.e., β. SD / β t Combining the mean t and standard deviation t of travel time for each road segment under different modes of transportation SD We can obtain the travel time budget B for each route segment for traveler i. This is used to obtain the K shortest path for each combination of travel modes in each region.

[0095] Step S3: Generate the initial selection set for regional travel.

[0096] Based on the origin and destination O of each route segment for traveler i i and D i Determine the set N of regional travel modes;

[0097] For any combination of travel modes n in a region, the K-shortest path algorithm is used to determine the initial set of selected paths K under that combination of travel modes, with travel time budget as the indicator. n ;

[0098] Combine the regional travel mode set N with the initial selected path set K n Perform joint generation of the initial selection set for regional travel.

[0099] Figure 3This invention provides a flowchart of a K-shortest path algorithm design. The K-shortest path algorithm uses a bidirectional sweep algorithm, with the travel time budget B as the route selection index to determine the initial set of selected paths K under different regional travel modes. n The specific steps are as follows:

[0100] Step 1 initializes the parameters. Set the start point O and end point D, set K=10, and create two queues Q. O and Q D Store the start point (O) and end point (D) separately, and initialize the distance array Dist. O and Dist D Record the shortest distance from the starting point (O) and the ending point (D) to each node respectively;

[0101] Step 2 involves alternating between forward and reverse searches. For example, Q... O and Q D If there are nodes, then take one node from one of the queues and expand it;

[0102] Step 3: Forward search. Starting from Q... O Extract node (u) from Dist and expand it, then update Dist. O For the neighboring node (v) of (u), if (Dist) O (u)+weight(u,v) <Dist O (v)), Update Dist O (v). If (v) is already in Q. O The middle part indicates that a path has been found and stored;

[0103] Step 4: Reverse search. From Q... D Extract node (v') from Dist and expand it, then update Dist. D For the neighboring node (u') of (v'), if (Dist) O (v')+weight(v',u') <Dist O (u')), update Dist O (u'). If (u') is already in Q. D The middle part indicates that a path has been found and stored;

[0104] Step 5: Path merging. In each expansion step, check Q. O Does the node in Q exist? D In the middle, or vice versa. If there is an intersection, it means that a path from (O) to (D) has been found. Store the found path and continue searching until K paths are found or the queue is empty.

[0105] Step S4: Calculate the traveler's tolerance time threshold.

[0106] The traveler tolerance time threshold function is constructed using a logarithmic function: T = α·ln(|t min |+1)+t min The parameter α = 0.526;

[0107] Based on the origin and destination O i and D i Determine the average travel time t of traveler i when choosing different regional combinations of travel modes n. in Extract the shortest travel time for traveler i when choosing different combinations of travel modes in different regions.

[0108] Given that traveler i selects a combination of regional travel modes n, the average travel time t for choosing travel route k is determined. ink Given that traveler i selects a combination of regional travel modes n, determine the shortest travel time for the chosen travel route k.

[0109] Calculate the travel time tolerance threshold for traveler i when choosing different regional combinations of travel modes n:

[0110] Calculate the travel time tolerance threshold for travel route k given that traveler i selects a combination of regional travel modes n:

[0111] Step S5: Determine the set of valid travel options within the region.

[0112] The travel time tolerance threshold T for regional combination travel modes in The average travel time t of traveler i choosing different regional combinations of travel modes n in Compare and select effective regional combinations of travel modes;

[0113] The travel time tolerance threshold T for choosing travel route k under the condition that traveler i selects a combination of regional travel modes n. ink Given that traveler i selects a combination of regional travel modes n, the average travel time t for choosing travel route k. ink Compare and filter the effective travel options;

[0114] Generate a joint choice set G = {G1; G2; ... G} of effective regional travel mode combinations and effective route selections. n′}, where G1={G 11 G 12 …G 1k′} represents the set of all valid choice paths k′ for the effective combination of regional travel modes n′.

[0115] Step S6: Calculate the probability of choosing an effective travel option within the region.

[0116] A latent class nested logit model is constructed, consisting of two parts: a latent classification model that categorizes travelers into different latent classes, and a mean-standard deviation (NL) model that combines travelers' choices of mode of transport and efficient routes. The NL model introduces the concept of a nest, i.e., dividing into multiple layers. It assumes that choices within the same nest are correlated, while choices in different nests are independent. The latent classification model automatically categorizes travelers by calculating the probability of a traveler belonging to each latent class based on personal attributes. There are two latent classes: based on the dummy variable values ​​in the personal attribute table, it can be determined whether traveler i belongs to latent class 1 and its probability; otherwise, it belongs to latent class 2.

[0117] The probability that traveler i belongs to potential category s is:

[0118]

[0119] V is =ASC is +∑η is ·X is

[0120] In the formula, V is For travelers i, this is a fixed term in the utility function when i belongs to the potential category s; For all potential class utilities and; ASC is X is a constant term when traveler i belongs to the potential category s; is Factors influencing whether traveler i belongs to potential category s include age, vehicle ownership, gender, income, and occupation. is For the corresponding parameters.

[0121] Table 3 Potential Category Classification and Corresponding Parameters of Influencing Factors

[0122]

[0123] The calculation result in this embodiment is that it belongs to potential category 1.

[0124] The utility function of the mean-standard deviation NL model is:

[0125]

[0126] In the formula, V igASC is a fixed term in the utility function when traveler i selects an effective choice limb g; ig For traveler i, the constant term is the effective choice limb g; t is the mean travel time for traveler i when choosing the effective choice limb g; t SD For traveler i, select the standard deviation of travel time under the effective choice limb; X ig Factors influencing traveler i's choice of effective limb g; η ig For the corresponding parameters;

[0127] Calculate the probability that traveler i belongs to potential category s, chooses the n′ combination of travel modes, and selects the effective travel path k′.

[0128]

[0129] In the formula, X isn′k′ Factors influencing traveler i's choice of the n′ regional combination travel mode and selection of the effective route k′, belonging to potential category s, include travel purpose, travel cost, travel distance, number of transfers, and familiarity with the road network. X is an influencing factor in selecting an effective route k′ given that traveler i belongs to potential category s and chooses regional combination travel mode n′. isn′ For travelers i belonging to potential category s, the factors influencing their choice of regional combination travel mode n′ are: η isn′k′ , η isn′ For the corresponding parameters.

[0130] Calculate the probability that traveler i chooses each valid option g:

[0131]

[0132] In the formula, V is For travelers i, this is a fixed term in the utility function when i belongs to the potential category s; For the utility of all potential categories; X is Factors influencing whether traveler i belongs to potential category s include age, vehicle ownership, gender, income, and occupation. is For the corresponding parameters; The fixed term of the utility function for selecting an effective choice path k′ under the condition that traveler i belongs to potential category s and chooses regional combination travel mode n′; V represents the utility sum of all valid alternative routes k′ given that traveler i belongs to potential category s and chooses a combination of regional travel modes n′; isn′ For traveler i belonging to potential category s, select the regional combination of travel mode n′ as a fixed term in the utility function; Let be the utility sum of all regional travel combination modes n′ for traveler i belonging to potential category s; LOGSUM is the composite utility term of the lower-level travel route selection limb for traveler i choosing the n′th effective regional travel combination mode. μ is the scaling coefficient connecting the upper and lower layers of the NL model.

[0133] Table 4. Parameter values ​​corresponding to influencing factors in the latent classification NL model.

[0134]

[0135] Step S7: Generate a personalized regional travel plan.

[0136] The output is a joint set of combined travel modes and valid routes in the region, sorted according to the selection probability of each valid option. The output includes combined travel mode, travel route, estimated time and cost, estimated departure time (corresponding to the expected arrival time as input) and estimated arrival probability ξ (corresponding to the expected departure probability ε as input).

[0137] Calculate β corresponding to the effective selection limb SD β t Value

[0138] Given the time budgets for each chosen route under different regional travel modes for traveler i, and β when traveler i belongs to different latent categories 1 and 2 as determined by the latent classification NL model, t The parameter values ​​are used to calculate β corresponding to the effective choice g of traveler i. SD β t Take values ​​and calculate β to satisfy the expected arrival probability ε. SD Data, combined with β from the model parameter calibration results t Calculate the corresponding β SD Used for calculating the probability of on-time arrival.

[0139] β of partially effective selected limbs SD β t The data is shown in the table below:

[0140] Table 5 shows the β values ​​of some effectively selected limbs g. SD β t data

[0141]

[0142] The travel time risk coefficient for a traveler choosing a valid option is obtained by weighting and summing the travel time risk coefficients for each potential category according to the probability of the traveler belonging to each potential category. Travel time risk coefficient calculation:

[0143] RRg =P 1i ·RR1+P 2i ·RR2

[0144] Among them, RR g RR1 represents the travel time risk coefficient for traveler i when choosing an effective option g; RR2 represents the travel time risk coefficient for traveler i belonging to potential category 1; P represents the travel time risk coefficient for traveler i belonging to potential category 2. 1i P 2i Let be the probability that traveler i belongs to potential category 1 or potential category 2.

[0145] The probability of on-time arrival for a traveler choosing a given valid option is calculated based on the travel time risk coefficient for that option. The traveler's probability of on-time arrival, ξ, is based on the coefficient β. t and the coefficient β that satisfies the expected arrival probability ε SD The calculated probability ξ of on-time arrival is as follows:

[0146]

[0147] In the formula, ξ g The probability of on-time arrival for traveler i when choosing an effective choice g is RR. g The travel time risk coefficient representing the effective choice limb g chosen by traveler i; β SD β represents the standard deviation coefficient of travel time for different options. t The mean travel time coefficients for different options are shown in Table 5.

[0148] Output the final, accurate, and reliable personalized travel plan recommendations. Filter the results where the traveler's on-time arrival probability ξ meets the expected on-time arrival probability ε, and the probability of selecting an effective option is greater than a threshold R. Sort the top three travel options by selection probability and output the following information: combined travel mode, travel route, estimated travel time, travel cost, on-time arrival probability, and recommended departure time.

[0149] Figure 4 , Figure 5 and Figure 6 These are the effective paths corresponding to recommended schemes one, two, and three provided in the embodiments of the present invention.

[0150] Table 6 Recommended Trip Results

[0151]

[0152] The corresponding recommended travel routes are as follows:

[0153] Table 7 Recommended Path

[0154]

[0155] In summary, the embodiments of the present invention perform potential classification of travelers' personal attributes, analyze the travel characteristics of potential categories, and construct a joint selection behavior model of shortest path selection and regional combination travel mode selection, so as to provide better personalized information services for regional travel.

[0156] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of one embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing the present invention.

[0157] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0158] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0159] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A personalized regional travel plan planning method considering travel time reliability, characterized in that, include: Initialize regional traffic supply information and regional traffic travel information, obtain the personal attributes and travel mode attributes of travelers, and calculate the travel time budget for each road segment for traveler i. Origin and destination of each route segment based on traveler i i and D i Given a set N of combined regional travel modes, use the K-shortest path algorithm to calculate the initial path selection set K for each combined regional travel mode. n ; Obtain the average travel time t of traveler i choosing different regional combinations of travel modes n. in Determine the shortest travel time for traveler i when choosing different regional combinations of travel modes n. And calculate its travel time tolerance threshold T in ; Obtain the average travel time t of traveler i choosing travel route k under the condition that the traveler i selects regional combination travel mode n. ink Given that traveler i selects a combination of regional travel modes n, determine the shortest travel time for choosing travel route k. And calculate its travel time tolerance threshold T ink ; T in With the t in Compare and filter effective regional combinations of travel modes; [the T] ink With the Compare and filter effective travel options to generate a set G of effective regional combinations of travel modes and effective options for effective travel options. Calculate the probability that traveler i belongs to potential category s, calculate the probability that traveler i chooses the n′ combination of travel modes and the effective travel route k′ when belonging to potential category s, and then calculate the probability that traveler i chooses each effective option g: Based on the probability of traveler i choosing each valid option g, sort and output the joint set of regional combination travel modes and valid selected routes, and output the personalized travel plan recommendation results for traveler i, including combination travel mode, travel route, estimated time and cost, estimated departure time and estimated arrival probability.

2. The method according to claim 1, characterized in that, The initialization of regional traffic supply information and regional traffic travel information, and the acquisition of travelers' personal attributes and travel mode attributes, include: Obtain multi-level rail transit operation plans and bus network operation plans for the loaded area, including the travel time distribution of urban roads and highways at all times of the day, the rail transit station entry and exit time distribution, the transfer time distribution within the same rail transit system and between different rail transit systems, as well as the travel cost (Cost) and the mean t and variance t of the travel time for each mode of transportation and each travel route. SD information; Obtain the traveler's personal attributes and travel mode attributes. Personal attributes include age, occupation, education, and monthly income. Travel mode attributes include the regional origin O. i Destination D in the region i Purpose of travel P i Information on expected arrival probability ε and estimated arrival time.

3. The method according to claim 2, characterized in that, The calculation of travel time budgets for each route segment for traveler i includes: Assuming that the travel time data of each traveler are independent, we fit the travel time data of each traveler to a normal distribution, and obtain the mean t and standard deviation t of the travel time for each road segment under different modes of transportation. SD ; Based on the above formula, the formula for calculating traveler i's travel time budget B can be derived as follows: In the formula, β t and β SD These are the coefficients of the mean travel time t and the standard deviation travel time t in the utility function of the mean-standard deviation NL model, respectively. SD The coefficient; Under the desired conditions, the mathematical relationship between the expected arrival time probability ε and the travel time budget B is as follows: t α Indicates the traveler's actual travel time; The formula can be transformed to obtain: RR represents the travel time risk coefficient, which indicates a traveler's attitude towards the trade-off between the mean travel time and the uncertainty of travel time. Based on the above formula, the RR (Return Rate) β of traveler i is calculated according to the expected arrival time probability ε. SD / β t Combining the mean t and standard deviation t of travel time for each road segment under different modes of transportation SD Calculate the travel time budget B for each segment of traveler i.

4. The method according to claim 3, characterized in that, The origin and destination O of each route segment based on traveler i i and D i Given a set N of combined regional travel modes, use the K-shortest path algorithm to calculate the initial path selection set K for each combined regional travel mode. n ,include: Origin and destination of each route segment based on traveler i i and D i Determine the set N of regional travel modes; For any regional combination of travel modes n in the set N of combined travel modes, the K-shortest path algorithm is used to determine the initial set of selected paths K under the combined travel modes in that region, with travel time budget as the indicator. n ; Combine the regional travel mode set N with the initial selected path set K n Perform joint generation of the initial selection set for regional travel.

5. The method according to claim 4, characterized in that, The method described above obtains the average travel time t of traveler i when choosing different regional combinations of travel modes n. in Determine the shortest travel time for traveler i when choosing different regional combinations of travel modes n. And calculate its travel time tolerance threshold T in ; Obtain the average travel time t of traveler i choosing travel route k under the condition that the traveler i selects regional combination travel mode n. ink Given that traveler i selects a combination of regional travel modes n, determine the shortest travel time for choosing travel route k. And calculate its travel time tolerance threshold T ink ,include: Based on the travel origin O of traveler i i and endpoint D i Determine the average travel time t of traveler i when choosing different regional combinations of travel modes n. in Extract the shortest travel time for traveler i when choosing different combinations of travel modes in different regions. Construct a traveler tolerance time threshold function: Given that traveler i selects a combination of regional travel modes n, the average travel time t for choosing travel route k is determined. ink Given that traveler i selects a combination of regional travel modes n, determine the shortest travel time for the chosen travel route k. Construct a traveler tolerance time threshold function:

6. The method according to claim 5, characterized in that, The T in With the t in Compare and filter effective regional combinations of travel modes; [the T] ink With the t ink By comparing and filtering effective travel options, a set G of effective choices for effective regional combinations of travel modes and effective routes is generated, including: The travel time tolerance threshold T for regional combination travel modes in The average travel time t of traveler i choosing different regional combinations of travel modes n in Compare and select effective regional combinations of travel modes; The travel time tolerance threshold T for choosing travel route k under the condition that traveler i selects a combination of regional travel modes n. ink Given that traveler i selects a combination of regional travel modes n, the average travel time t for choosing travel route k. ink Compare and filter the effective travel options; Generate a joint choice set G = {G1; G2; ... G} of effective regional travel mode combinations and effective route selections. n′ }, where G1={G 11 G 12 …G 1k′ } represents the set of all valid choice paths k′ for the effective regional combination travel mode n′.

7. The method according to claim 6, characterized in that, The calculation of the probability that traveler i belongs to potential category s, the calculation of the probability that traveler i chooses the n′ combination of travel modes and the effective travel path k′ when belonging to potential category s, and the calculation of the probability that traveler i chooses each effective option g, includes: Construct a latent classification model for travelers, and use this model to divide travelers into different latent categories. The probability that traveler i belongs to latent category s is: V is =ASC is +∑η is ·X is In the formula, V is For travelers i, this is a fixed term in the utility function when i belongs to the potential category s; For all potential class utilities and; ASC is X is a constant term when traveler i belongs to the potential category s; is Factors influencing whether traveler i belongs to potential category s include age, vehicle ownership, gender, income, and occupation. is For the corresponding parameters; The utility function of the mean-standard deviation NL model is: In the formula, V ig ASC is a fixed term in the utility function when traveler i selects an effective choice limb g; ig For traveler i, the constant term is the effective choice limb g; t is the mean travel time for traveler i when choosing the effective choice limb g; t SD For traveler i, select the standard deviation of travel time under the effective choice limb; X ig Factors influencing traveler i's choice of effective limb g; η ig For the corresponding parameters; The probability of traveler i choosing the n′ combination of travel modes and the effective travel path k′ is calculated based on the utility function of the mean-standard deviation NL model. In the formula, X isn′k′ Factors influencing traveler i's choice of the n′ regional combination travel mode and selection of the effective route k′, belonging to potential category s, include travel purpose, travel cost, travel distance, number of transfers, and familiarity with the road network. X is an influencing factor in selecting an effective route k′ given that traveler i belongs to potential category s and chooses regional combination travel mode n′. isn′ For travelers i belonging to potential category s, the factors influencing their choice of regional combination travel mode n′ are: η isn′k′ , η isn′ For the corresponding parameters; Calculate the probability that traveler i chooses each valid option g: In the formula, LOGSUM represents the combined utility term of traveler i choosing the n′th effective regional combination travel mode and the lower-level travel route selection limb, and μ is the scaling coefficient connecting the upper and lower layers of the NL model.

8. The method according to claim 7, characterized in that, The process of sorting and outputting a joint set of regional combined travel modes and effective selected routes based on the probability of traveler i choosing each effective option g, and outputting a personalized travel plan recommendation result for traveler i, including combined travel modes, travel routes, estimated travel time and cost, estimated departure time, and estimated arrival probability, including: The output is a joint set of regional travel modes and effective routes, sorted according to the selection probabilities of each effective option. Based on the time budget of each route under each regional travel mode, and combined with the β value obtained from the latent classification NL model indicating whether the traveler belongs to different latent categories... t The parameter values ​​are used to calculate β corresponding to the effective choice made by traveler i. SD β t Values; The travel time risk coefficients of travelers belonging to each potential category and choosing a certain effective option are weighted and summed according to the probability of the traveler belonging to each potential category to obtain the travel time risk coefficient of the traveler choosing a certain effective option; the on-time arrival probability of the traveler choosing a certain effective option is calculated based on the travel time risk coefficient of the traveler choosing a certain effective option. The system filters travel options whose on-time arrival probability ξ satisfies the expected on-time arrival probability ε and whose effective selection probability is greater than the threshold R. The options are sorted according to their selection probability, and the first set number of travel options are output as the personalized travel option recommendation results for traveler i. These options include the combination of travel modes, travel route, estimated time and cost, estimated departure time and estimated arrival probability.

Images