CN116628527B - Design method and system for integrated travel strategy - Google Patents

Design method and system for integrated travel strategy Download PDF

Info

Publication number
CN116628527B
CN116628527B CN202310910715.XA CN202310910715A CN116628527B CN 116628527 B CN116628527 B CN 116628527B CN 202310910715 A CN202310910715 A CN 202310910715A CN 116628527 B CN116628527 B CN 116628527B
Authority
CN
China
Prior art keywords
trip
data
travel
clustering
integrated
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202310910715.XA
Other languages
Chinese (zh)
Other versions
CN116628527A (en
Inventor
孙轶琳
赵志健
何艳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang University ZJU
Original Assignee
Zhejiang University ZJU
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhejiang University ZJU filed Critical Zhejiang University ZJU
Priority to CN202310910715.XA priority Critical patent/CN116628527B/en
Publication of CN116628527A publication Critical patent/CN116628527A/en
Application granted granted Critical
Publication of CN116628527B publication Critical patent/CN116628527B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Abstract

The application provides a design method of an integrated travel strategy, which comprises the following steps: acquiring RP trip investigation data, wherein the RP trip investigation data comprises a trip starting point, a trip ending point, trip duration and trip modes; performing center point clustering on the RP travel investigation data to obtain a clustering result, wherein the clustering result comprises the collocation range of different types of users on any type of traffic mode; according to the clustering result, a D-effect design is adopted, a numerical level is selected in the collocation range of each type of user to any type of traffic mode to serve as the quota level of the type of traffic mode, so that a selection situation is generated, the selection situation comprises at least two integrated travel strategies, and any integrated travel strategy comprises the use quota of at least two types of traffic tools. The integrated travel strategy design method provided by the application can solve the problem that the accurate identification cannot be carried out aiming at the requirements of users on the time span existing in the related technology.

Description

Design method and system for integrated travel strategy
Technical Field
The application relates to the technical field of traffic, in particular to a design method and system for an integrated travel strategy.
Background
With the development of Information and Communication Technology (ICT), travel, namely, the appearance of service ((Mobility as a Service, maaS) provides a new strategy for travel management, integrated and integrated transportation travel is realized by integrating traffic resources through a digital platform, so that travel service from 'door' to 'door' is provided for travelers, the idealized MaaS platform breaks through barriers of various transportation modes such as trains, subways, buses, taxis, shared automobiles, shared bicycles and the like, and the travel service in a period of time can be provided for users by inquiring, ordering, paying, travel reservation and the like of all transportation modes through one mobile phone application.
However, the existing travel service in the related art aims at providing a single travel service, and cannot accurately identify the needs of the user in a time span. For example, although the existing travel service can provide different travel schemes for a single travel of a user, in a time span, such as a week or a month, the existing technology cannot identify the user requirement in the time, and cannot recommend a more reasonable travel strategy for the user.
Therefore, a technical scheme is needed to solve the problem that the time span in the related technology cannot be accurately identified according to the needs of the user, and provide MaaS-type travel service for the user.
Disclosure of Invention
The application aims to provide a technical scheme for solving the problem that the time span in the related technology cannot be accurately identified according to the requirements of users.
Based on the problems, the application provides a design method of an integrated travel strategy, which comprises the following steps:
acquiring RP trip investigation data, wherein the RP trip investigation data comprises a trip starting point, a trip ending point, trip duration and trip modes;
performing center point clustering on the RP travel investigation data to obtain a clustering result, wherein the clustering result comprises the collocation range of different types of users on any type of traffic mode;
according to the clustering result, a D-effect design is adopted, a numerical level is selected in the collocation range of each type of user to any type of traffic mode to serve as the quota level of the type of traffic mode, so that a selection situation is generated, the selection situation comprises at least two integrated travel strategies, and any integrated travel strategy comprises the use quota of at least two types of traffic tools.
Further, when the central point clustering is performed, clustering is performed on RP trip investigation data by adopting a K-medoids clustering algorithm.
Further, clustering the RP trip survey data by adopting a K-means clustering algorithm comprises the following steps:
preprocessing RP trip investigation data to construct a data set, wherein the data set comprises a plurality of pieces of attribute data of each traveler, the attribute data is used as coordinates of sample points, and the attribute data of the traveler comprises the number of public transportation trips, the distance of a car trip and the duration of a bicycle trip of the traveler in a preset time range;
random selectionkThe sample points are used as center points, and Euclidean distance from the rest sample points to any center point is calculated;
according to the calculated Euclidean distance, according to the principle of the nearest center, the rest sample points are distributed to the categories represented by the center points, so that initial clustering is realized;
calculating a clustering center again, traversing each sample point in the category, calculating Euclidean distance from the sample point to the rest sample points in the category, selecting a group corresponding to the minimum Euclidean distance, and taking the group as a new center point;
and carrying out clustering division again according to the new center points until all the center points are not changed any more or the maximum iteration times are reached, so as to obtain the clustering results of the current k categories.
Further, the euclidean distance from any sample point to a certain center point is expressed by the following formula:
；
wherein:d i is the firstiEuclidean distance of a sample point to the center point,x ij is the data setiLine (1)jThe attribute data of the column,x mj is the firstmFirst of the center pointsjColumn attribute data.
Further, when the RP trip investigation data is clustered by adopting a K-medoids clustering algorithm, the number of clustered categories is determined by the following formula:
；
in the method, in the process of the application,C k is the firstkThe data of the category(s),sis thatC k Is used for the data of the data in the database,o k is thatC k Is taken from the centroid of (1)SSETurning point of downward trendKThe value is the number of categories of the cluster, and each category in the cluster is used as the user of the same type.
Further, the method further comprises:
designing at least two subsidy schemes of at least one type of traffic mode, adopting D-effect design according to the clustering result and the subsidy schemes, and selecting a numerical level as a limit level of the type of traffic mode in the collocation range of each type of user to any type of traffic mode to generate a selection situation, wherein the selection situation comprises at least two integrated travel strategies.
Further, the integrated travel policy includes at least one or more of the following types of vehicle usage credits:
the number of times public transportation is available, the total available mileage of the net-bound vehicle, the unit price of the net-bound vehicle mileage, and the available time of the leased vehicle.
Further, the method further comprises:
collecting SP investigation data, obtaining SP investigation data, and preprocessing the SP investigation data, wherein the SP investigation data comprises: traveler personal attribute data, travel behavior attribute data and integrated travel strategy attribute data;
constructing a Mixed Logit model, inputting the preprocessed SP investigation data into the Mixed Logit model, and carrying out parameter estimation on the Mixed Logit model by using a maximum simulation likelihood estimation algorithm;
and evaluating the parameter estimation result, and judging whether the integrated trip strategy needs to be optimized according to the evaluation result.
Further, a utility function is constructed, the utility function being represented by the following formula:
；
in the method, in the process of the application,V ntj indicating the utility of the observation,nthe person who is going out is indicated,tthe selection context is indicated and,jan integrated travel policy option is represented,Zrepresents a matrix of individual attribute variables of the traveler,Yrepresents a travel behavior attribute matrix,Xrepresenting a matrix of attributes associated with the integrated travel policy,γ、φandβrepresenting the parameters to be estimated and,ε ntj representing error items, obeying independent identical Gumbel distribution;
according to the utility function, calculating the selection probability of the interviewee on the integrated travel strategy, wherein the selection probability is expressed by the following formula:
；
solving a log-likelihood function of the Mixed Logit model, wherein the log-likelihood function is expressed by the following formula:
；
wherein if a context is being selectedtLower optioniInterviewee personnSelect thenEqual to 1, otherwise->Equal to 0;
solving the log-likelihood function to obtain the estimated values of parameters of the usage amount of different types of traffic tools in the integrated trip strategy and the corresponding estimated valuesPValue whenPWhen the value is larger than the preset threshold value, judging that the estimated value of the parameter is obvious in the confidence interval, whenPWhen the value is smaller than or equal to a preset threshold value, judging that the integrated trip strategy is needed to be integrated with the integrated trip strategyPThe usage amount of the traffic tool corresponding to the value is optimized.
The application also provides an integrated travel strategy design system, which adopts the design method of the integrated travel strategy.
In summary, the application provides a design method of an integrated travel strategy, which comprises the steps of firstly clustering according to RP travel survey data, adopting D-effect design according to clustering results, generating the integrated travel strategy for different types of users, and determining which type of travel strategy to use according to the self preference, the demand, the behavior habit and other factors of the users, thereby meeting the travel demands of different users on a certain time span, reducing the travel cost, guiding public transportation travel, reducing carbon emission and other beneficial effects, promoting sustainable travel and assisting carbon emission reduction in the transportation field.
Drawings
FIG. 1 is a flow chart of an integrated travel strategy design method provided by an embodiment of the present application;
FIG. 2 is a schematic diagram of a selection scenario according to an embodiment of the present application;
FIG. 3 is a flowchart of performing center point clustering on RP trip survey data by adopting a K-means clustering algorithm, which is provided by the embodiment of the application;
FIG. 4 is a schematic diagram illustrating a selection scenario according to another embodiment of the present application;
fig. 5 is a flowchart of an integrated trip strategy design method according to another embodiment of the present application.
Detailed Description
The present application will be described in detail below with reference to the specific embodiments shown in the drawings, but these embodiments are not limited to the present application, and structural, method, or functional modifications made by those skilled in the art based on these embodiments are included in the scope of the present application.
As shown in fig. 1, an embodiment of the present application provides a design method for an integrated trip policy, including the following steps:
s1, acquiring RP (Revealed preference, behavior investigation) travel investigation data, wherein the RP travel investigation data comprises a travel starting point, a travel end point, travel duration and a travel mode;
s2, performing center point clustering on the RP trip investigation data to obtain a clustering result, wherein the clustering result comprises the collocation range of different types of users on any type of traffic mode;
s3, adopting a D-effect design according to a clustering result, and selecting a numerical level as a limit level of any type of traffic mode in the collocation range of each type of user to the type of traffic mode to generate a selection situation, wherein the selection situation comprises at least two integrated travel strategies, and any integrated travel strategy comprises the use limit of at least two types of traffic tools.
As an optional implementation manner, in an embodiment of the present application, different usage amounts may be set for different types of vehicles, and the integrated trip policy includes at least one or more of the following types of vehicle usage amounts: the number of times public transportation is available, the total available mileage of the net-bound vehicle, the unit price of the net-bound vehicle mileage, and the available time of the leased vehicle.
As shown in fig. 2, which schematically illustrates a selection scenario for the design of an embodiment of the present application. In the embodiment of the application, the designed selection situation comprises three integrated travel strategies, and each travel strategy comprises the use limit of different types of vehicles. For example, policy a is configured such that the amount of payment required by the user to purchase policy a is 230 yuan. After purchasing policy a, the user may optionally govern the usage amount of various vehicles contained in policy a during the validity period of policy a. That is, during the expiration date of the policy a, the purchaser can take the public transportation means twice at will according to the policy content, the total mileage of the available net car is about 110 km, the bicycle can be used at will within 1 hour, and no extra cost is generated when the transportation means is used within the use limit of any transportation means. Regarding the effective period of the strategy, the effective period of the strategy can be set according to actual requirements, for example, the effective period of the strategy can be set to be one week or one month, and the method can be suitable for the travel demands of different types of people, such as work commute, social medical treatment, leisure travel, and the like, so that a traveler can enjoy the multi-mode travel service on one application, the trouble of frequent switching of application and payment mode disputes is solved, and travel convenience and happiness are improved.
According to the description, the design method for the integrated travel strategy is provided, firstly, clustering is carried out according to RP travel survey data, then, D-effect design is adopted according to clustering results, integrated travel strategies are generated for different types of users, and the users can decide which type of travel strategy to purchase according to the self preference, the demand, the behavior habit and other factors, so that the travel demands of different users on a certain time span can be met, the travel cost is reduced, the beneficial effects of guiding public transportation travel, reducing carbon emission and the like are achieved, sustainable travel can be promoted, and carbon emission reduction in the power-assisted transportation field is facilitated.
As an alternative implementation, in step S1, RP travel survey data may be acquired from a travel database of the target city. In the embodiment of the application, resident trip investigation data in a week are acquired from a trip database of a target city. The system mainly comprises attributes such as a travel starting point, a travel ending point, a travel mode, travel time and the like, and sample data are shown in table 1.
TABLE 1
As shown in fig. 3, in step S2, the RP travel survey data may be clustered by using a K-means clustering algorithm. Specifically, the method comprises the following steps:
s21, preprocessing RP trip investigation data to construct a data set, wherein the data set comprises a plurality of items of attribute data of each traveler, and the attribute data is used as coordinates of sample points. The attribute data of the traveler comprises the number of public transportation trips, the distance of travel of the car and the travel time of the bicycle, wherein the number of public transportation trips, the distance of travel of the car and the travel time of the bicycle occur in a preset time range;
s22, randomly selectingkThe sample points are used as center points, and Euclidean distance from the rest sample points to any center point is calculated;
s23, according to the calculated Euclidean distance and the principle of nearby centers, distributing the rest sample points to the categories represented by the center points to realize initial clustering;
s24, recalculating a clustering center, traversing each sample point in the category, calculating Euclidean distance from the sample point to the rest sample points in the category, and selecting a group corresponding to the minimum Euclidean distance as a new center point;
s25, carrying out clustering division again according to the new center points until all the center points are not changed or the maximum iteration number is reached, so as to obtain the clustering results of the current k categories.
As an optional implementation manner, in step S22, K-means clustering is performed based on attribute data of a traveler in a target city, K sample points are randomly selected as center points, in the embodiment of the present application, the value range of K is set to be greater than 1 and less than or equal to 10, the euclidean distance between the remaining sample points and any center point is calculated, and the euclidean distance can be obtained by calculating according to the following formula (1):
（1）
wherein:d i is the firstiEuclidean distance of the sample points to the merodids;x ij is the data setiLine (1)jThe number of columns is determined by the number of columns,x mj is the firstmFirst of the center pointsjColumn data; the rows of the dataset may represent travelers and the columns may represent extracted attribute data.
As an optional implementation manner, when the RP travel survey data is clustered by adopting a K-means clustering algorithm, the number K of clustered categories is determined by the following formula (2):
（2）
in the method, in the process of the application,C k is the firstkThe data of the category(s),sis thatC k Is used for the data of the data in the database,o k is thatC k Is a centroid of (c).
Taking outSSETurning point of downward trendKThe value is the number of categories of the cluster, and each category in the cluster is used as the user of the same type.
In step S25, clustering is performed according to the optimal value of k, and a final multi-mode combined clustering result is obtained, as shown in table 2, which exemplarily shows clustering results that can be further applied.
TABLE 2
As shown in Table 2, in the embodiment of the application, RP trip survey data is centrally clustered, and travelers are classified into 3 types of users, namely public transportation users, automobile users and multimode users. In the embodiment of the application, the RP trip survey data is resident trip survey data in a week, which is acquired from a trip database of a target city. And clustering RP travel investigation data, wherein the matching range of the times of riding public transportation in a week is within [10, 14], the matching range of the mileage of riding the car is within [0km,45km ], and the users using the bicycle duration within [4h,7h ] are classified as public transportation users. And classifying the users with the matching range of riding public transportation times within a week within [0,5] and the matching range of riding car mileage within [90km,150km ] and the matching range of using bicycle duration within [0h,3h ] as car users. Users with the matching range of riding public transportation times within a week within [6,9] and the matching range of riding car mileage within [50km,100km ] and the matching range of using bicycle duration within [2h,5h ] are classified as multimode users. The number of public transportation is an integer in the section.
In step S3, an integrated trip strategy is generated for different types of users by adopting a D-effect design according to the clustering result. The D-effect design is a design algorithm for discrete selection experiments. The code can be written by using Ngene software, the discrete selection experimental situation design is carried out by adopting a D-effect design method, the number of generated situations can be set according to actual requirements, and generally, 4-8 situations are needed.
The step S3 specifically comprises the following steps:
s31, under the condition of meeting the requirements of horizontal equalization, orthogonality and minimum overlapping, selecting an applied numerical value level in a numerical value level collocation range according to a clustering result to generate an initial design matrixX. The design matrix X represents a selection scenario designed according to the embodiment of the present application.
Level equalization requires equal opportunities for attribute levels to appear in the dataset, orthogonality requires that combinations of attribute levels exist in a particular correlation pattern, and minimal level overlap pursues minimizing the repetition probability of attribute levels in the same set of experimental scenarios to ensure validity of attribute importance estimates.
Specifically, according to the clustering result in table 2, the numerical level collocation range among the number of public transportation, the number of kilometers of the car and the number of hours of the bicycle can be obtained, and the numerical level applied in the collocation range can be selected when the travel strategy design is carried out. For example, an integer in the range of public transportation times is selected, an integer multiple of a preset kilometer is selected in the range of kilometer of the car, and a multiple of a preset time in the range of hour of the bicycle is selected. For the convenience of calculation, in the embodiment of the present application, the preset kilometers are set to 5 or 10, and the preset time is set to 0.5.
As an optional implementation manner, in the embodiment of the application, a taxi discount (10%, 20%) can be added to the integrated trip policy. As shown in fig. 4, a travel policy option provided by an embodiment of the present application in a selection context is shown. Each selection context may include two custom travel policies, one pay-per-use travel policy, and no-subscription options.
S32, according to the design matrixXAnd determining a progressive variance-covariance (AVC) matrix for predicting the current design context with respect to some prior probability distribution information for parameter estimationXAnd the next trip modes are horizontally matched with parameter estimation errors generated by the data obtained by design.
As an alternative implementation, the efficiency of the design may be based onPerforming weighing; to be used forAnd (3) solving the problem by adopting a Modified Federov algorithm with the minimum as a target, and iteratively searching for an approximate optimal feasible design. />The calculation method is as follows:
（3）
wherein:Pis the number of parameters to be estimated,a progressive variance-covariance (AVC) matrix representing an individual interviewee, the AVC matrix being a P-x-P matrix, which is a design matrixXPrior probability distributionInformation->Function of->Can be obtained from similar studies or pre-surveys in order to make +.>Independent of the size of the problem, it is normalized to 1 +.PTo the power.
According to the description, the integrated travel strategy design method provided by the embodiment of the application can generate integrated travel strategies for different types of users, and the users can decide which type of travel strategy to purchase according to the self preference, the demand, the behavior habit and other factors, so that the travel demands of different users on a certain time span can be met, the travel cost is reduced, the public transportation travel can be guided, the carbon emission is reduced and other beneficial effects can be achieved, the sustainable travel can be promoted, and the carbon emission in the power-assisted transportation field can be reduced.
As shown in fig. 5, as an optional implementation manner, the integrated trip policy design method provided by the embodiment of the present application further includes:
s4, collecting SP investigation data, obtaining SP investigation data, preprocessing the SP investigation data, wherein the SP investigation data comprises the following steps: traveler personal attribute data, travel behavior attribute data and integrated travel strategy attribute data;
s5, constructing a Mixed Logit model, inputting the preprocessed SP investigation data into the Mixed Logit model, and carrying out parameter estimation on the Mixed Logit model by using a maximum simulation likelihood estimation algorithm;
and S6, evaluating the parameter estimation result, and judging whether the integrated travel strategy needs to be optimized according to the evaluation result.
As an optional implementation manner, in step S4, the traveler personal attribute data includes: gender, age, academic, profession, home structure, home car possession, home month income, and major travel patterns in daily and travel.
In the embodiment of the application, the age attribute data of the traveler is divided according to the following age ranges, wherein the age ranges comprise: under 18 years old, 18-30 years old, 31-50 years old, 51-65 years old and over 65 years old. The traveler's academic is classified into 4 categories, including: university specialty/professional technical colleges, university family, major and above. The profession of the traveler is classified into at least 6 categories, including: public officers/enterprises and institutions personnel, service personnel, individual operators, students, no business (including housewives, retirees, etc.), and others. The total household monthly revenue is classified into 5 classes: 5000 yuan below, 5000-9999 yuan, 10000-29999 yuan, 30000-49999 yuan, 50000 yuan and above. The home car possession was divided into 4 classes: 0, 1, 2, 3 and above.
The travel behavior attribute data in the SP survey data comprises: the trip data of the next week (including trip mode, trip distance and the like, and the main trip mode in daily life and the use frequency of each mode mainly comprise the data of 9 trip modes, namely private cars, buses, subways, taxi/network buses, carpools, shared cars, bicycles, walking and the like).
Integrating travel strategy attribute data in SP investigation data: including number of public traffic and network taxi mileage enjoyable, network taxi mileage price, taxi discount, sharing of the number of hours of a single vehicle, and trip policy price.
In step S4, the collected data needs to be preprocessed, and the method includes: and eliminating the data with the time of filling in the questionnaire less than a certain value, eliminating the data with the number of family members not conforming to the family structure, and eliminating the data with smaller variance of the selection situation. By the method, abnormal data in the SP investigation data can be initially removed, and interference to an evaluation result is avoided.
In step S5, the parameter estimation of the Mixed Logit model comprises the following steps:
constructing a utility function, the utility function being represented by the following formula (4):
(4)
in the method, in the process of the application,V ntj indicating the utility of the observation,nthe person who is going out is indicated,tthe selection context is indicated and,jan integrated travel policy option is represented,Zrepresents a matrix of individual attribute variables of the traveler,Yrepresents a travel behavior attribute matrix,Xrepresenting a matrix of attributes associated with the integrated travel policy,γ、φandβrepresenting the parameters to be estimated and,ε ntj representing error terms, subject to independent identical gummel distributions.
The selection situation is generated by performing discrete selection of experimental situation design by adopting a D-effect design method, and the number of the selection situations can be set according to actual requirements, and generally 4-8 situations are needed. The integrated trip strategy option is to indicate the pedestriannThe result of the selection among travel policy a, travel policy B, pay-as-you-go and unsubscribe in fig. 2.
As an alternative implementation, parts may be combinedSetting as random parameters to capture the preference heterogeneity of travelers for integrated travel strategies, ++>The distribution form of (2) can be represented by the following formula (5):
(5)
in the method, in the process of the application,representation and attributesk(certain Integrated travel policy Property) related parameters +.>Sample mean value of>Is a normal distribution random term with an average value of 0 and a standard deviation of 1,/>is->Is a standard deviation of the distribution of (c).
According to the utility function of the formula (4), calculating the selection probability of the interviewee on the integrated travel strategy, wherein the selection probability is represented by the following formula (6):
(6)
the SP survey data is converted into a format that can be computationally analyzed using Nlogit software, including efficient encoding of the classification variables. To better study the heterogeneity of SP investigated pseudo-panel data with interviewees, mixed Logit model parameter estimation was performed based on a maximum analog likelihood estimation method, using a Halton sampling method to sample the probability density function 1000 times and using the analog probability mean as an approximate solution of the integral.
Solving a log-likelihood function of the Mixed Logit model, wherein the log-likelihood function is represented by the following formula (7):
(7)
wherein if a context is being selectedtTrip policy optioniInterviewee personnSelect thenEqual to 1, otherwiseEqual to 0.
Solving the log likelihood function to obtain the fitting degree R of the model 2 Parameters corresponding to personal attributes, travel behavior attributes and integrated travel strategy attributesγ、φAndβcorresponding to each parameterPValues.
In step S6, the parameter estimation result is evaluated, whenPWhen the value is larger than the preset threshold value, judging that the estimated value of the parameter is obvious in the confidence interval, whenPWhen the value is smaller than or equal to a preset threshold value, judging that the integrated trip strategy is needed to be integrated with the integrated trip strategyPThe usage amount of the traffic tool corresponding to the value is optimized.
Specifically, the above steps S5 to S6 may be written and run in the nlogic to obtain the parameter estimation result of the Mixed logic model. Evaluating the parameter estimation result of the model, firstly checking the fitting degree R of the model 2 The value of 0.2 and above can be considered that the fitting effect of the model to SP investigation data is good, which indicates the effectiveness of influence factor selection and the interpretation degree of the model.
Checking the estimated parameters of each influence factor, wherein the positive and negative of the numerical values respectively represent the positive and negative influence on the selection of the corresponding MaaS trip strategy byPThe value can be seen as a significance of the effect,P<0.1、P<0.05、P<0.01 indicates significant at 90%, 95%, 99% confidence levels, respectively.
The application potential of the current MaaS can be obtained through the parameter estimation result of the model, the target crowd image of the MaaS and the preference of the travel mode contained in the travel strategy are identified, the payment willingness of the traveler to the MaaS travel strategy is calculated, and some policy revenues and theoretical supports can be provided for related planners and operators in the aspects of optimizing the travel mode contained in the travel strategy, the level and pricing of the travel mode.
As an optional implementation manner, the embodiment of the application further provides an integrated travel strategy design system, and the system adopts the integrated travel strategy design method provided by the embodiment of the application.
The above disclosure is illustrative of the preferred embodiments of the present application, but it should not be construed as limiting the scope of the application as will be understood by those skilled in the art: changes, modifications, substitutions, combinations, and simplifications may be made without departing from the spirit and scope of the application and the appended claims, and equivalents may be substituted and still fall within the scope of the application.

Claims (7)

1. The design method of the integrated travel strategy is characterized by comprising the following steps of:
acquiring RP trip investigation data, wherein the RP trip investigation data comprises a trip starting point, a trip ending point, a trip duration and a trip mode;
performing center point clustering on the RP travel investigation data to obtain a clustering result, wherein the clustering result comprises the collocation range of different types of users on any type of traffic mode;
according to the clustering result, adopting a D-effect design, and selecting a numerical level as a limit level of any type of traffic mode in the collocation range of each type of user to any type of traffic mode to generate a selection situation, wherein the selection situation comprises at least two integrated travel strategies, any one of the integrated travel strategies comprises the use limit of at least two types of traffic tools, and the integrated travel strategies at least comprises the use limit of the following types of traffic tools: the available times of public transportation, the total available mileage of the network about vehicle, the unit price of the network about vehicle mileage and the available time of the leased transportation;
when the central point clustering is performed, clustering the RP trip investigation data by adopting a K-medoids clustering algorithm;
clustering the RP trip survey data by adopting a K-medoids clustering algorithm comprises the following steps:
preprocessing the RP trip investigation data to construct a data set, wherein the data set comprises a plurality of items of attribute data of each traveler, the attribute data is used as coordinates of sample points, and the attribute data of each traveler comprises the number of public transportation trips, the distance of a car trip and the duration of a bicycle trip of the traveler in a preset time range;
random selectionkThe sample points are used as center points, and Euclidean distance from the rest sample points to any center point is calculated;
according to the calculated Euclidean distance, according to the principle of the nearest center, the rest sample points are distributed to the categories represented by the center points, so that initial clustering is realized;
calculating a clustering center again, traversing each sample point in the category, calculating Euclidean distance from the sample point to the rest sample points in the category, and selecting a group corresponding to the minimum Euclidean distance as a new center point;
and carrying out clustering division again according to the new center points until all the center points are not changed any more or the maximum iteration times are reached, so as to obtain the clustering results of the current k categories.
2. The method for designing an integrated travel strategy according to claim 1, wherein,
the euclidean distance of any one of the sample points to a certain of the center points is expressed by the following formula:
wherein:d i is the firstiEuclidean distance of a sample point to the center point,x ij to the data setiLine (1)jThe attribute data of the column,x mj is the firstmFirst of the center pointsjColumn attribute data.
3. The method for designing an integrated travel strategy according to claim 1, wherein,
when the RP trip survey data is clustered by adopting a K-medoids clustering algorithm, the number of clustered categories is determined by the following formula:
in the method, in the process of the application,C k is the firstkThe data of the category(s),sis thatC k Is used for the data of the data in the database,o k is thatC k Is taken from the centroid of (1)SSETurning point of downward trendKThe value is the number of categories of the cluster, and each category in the cluster is used as the user of the same type.
4. The method for designing an integrated travel strategy according to claim 1, further comprising:
designing at least two subsidy schemes of at least one type of traffic mode, adopting D-effect design according to the clustering result and the subsidy schemes, and selecting a numerical level as the limit level of the type of traffic mode in the collocation range of each type of traffic mode by each type of user to generate a selection situation, wherein the selection situation comprises at least two integrated travel strategies.
5. The method for designing an integrated travel strategy according to claim 1, further comprising:
collecting SP investigation data, obtaining SP investigation data, and preprocessing the SP investigation data, wherein the SP investigation data comprises the following steps: traveler personal attribute data, travel behavior attribute data and integrated travel strategy attribute data;
constructing a Mixed Logit model, inputting the preprocessed SP investigation data into the Mixed Logit model, and carrying out parameter estimation on the Mixed Logit model by using a maximum simulation likelihood estimation algorithm;
and evaluating the parameter estimation result, and judging whether the integrated travel strategy needs to be optimized according to the evaluation result.
6. The method for designing an integrated travel strategy according to claim 5, wherein,
constructing a utility function, the utility function being represented by the following formula:
in the method, in the process of the application,V ntj indicating the utility of the observation,nthe person who is going out is indicated,tthe selection context is indicated and,jan integrated travel policy option is represented,Zrepresents a matrix of individual attribute variables of the traveler,Ythe representation showsA matrix of row behavior attributes,Xrepresenting a matrix of attributes associated with the integrated travel policy,γ、φandβrepresenting the parameters to be estimated and,ε ntj representing error items, obeying independent identical Gumbel distribution;
according to the utility function, calculating the selection probability of the interviewee on the integrated travel strategy, wherein the selection probability is expressed by the following formula:
solving a log-likelihood function of the Mixed logic model, the log-likelihood function being represented by the following formula:
wherein if a context is being selectedtLower optioniInterviewee personnSelect thenEqual to 1, otherwise->Equal to 0;
solving the log likelihood function to obtain parameter estimation values of the use limits of different types of traffic tools in the integrated trip strategy and corresponding parametersPValue when saidPWhen the value is larger than a preset threshold value, judging that the parameter estimation value is obvious in a confidence interval, and when the parameter estimation value is larger than a preset threshold valuePWhen the value is smaller than or equal to the preset threshold value, judging that the integrated trip strategy is required to be matched with the integrated trip strategyPThe usage amount of the traffic tool corresponding to the value is optimized.
7. An integrated travel strategy design system, characterized in that the system adopts the design method of the integrated travel strategy as claimed in any one of claims 1 to 6.
CN202310910715.XA 2023-07-24 2023-07-24 Design method and system for integrated travel strategy Active CN116628527B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310910715.XA CN116628527B (en) 2023-07-24 2023-07-24 Design method and system for integrated travel strategy

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310910715.XA CN116628527B (en) 2023-07-24 2023-07-24 Design method and system for integrated travel strategy

Publications (2)

Publication Number Publication Date
CN116628527A CN116628527A (en) 2023-08-22
CN116628527B true CN116628527B (en) 2023-11-10

Family

ID=87603028

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310910715.XA Active CN116628527B (en) 2023-07-24 2023-07-24 Design method and system for integrated travel strategy

Country Status (1)

Country Link
CN (1) CN116628527B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108564396A (en) * 2018-03-23 2018-09-21 同济大学 Multimode traffic trip questionnaire survey design method based on D-error design effectivelies
CN110390415A (en) * 2018-04-18 2019-10-29 北京嘀嘀无限科技发展有限公司 A kind of method and system carrying out trip mode recommendation based on user's trip big data
CN111737601A (en) * 2020-06-08 2020-10-02 北京奇虎科技有限公司 Method, device and equipment for recommending travel strategy and storage medium
CN112632374A (en) * 2020-12-18 2021-04-09 东南大学 Resident travel mode selection analysis method considering customized bus
CN113269358A (en) * 2021-05-19 2021-08-17 兆边(上海)科技有限公司 Planning method based on multi-mode integrated travel
CN113780808A (en) * 2021-09-10 2021-12-10 西南交通大学 Vehicle service attribute decision optimization method based on flexible bus connection system line
CN115049217A (en) * 2022-05-17 2022-09-13 中国平安财产保险股份有限公司 Method and device for generating travel strategy, computer equipment and storage medium
CN115412857A (en) * 2022-08-24 2022-11-29 浙江大学 Resident travel information prediction method
WO2023273292A1 (en) * 2021-06-30 2023-01-05 深圳市城市交通规划设计研究中心股份有限公司 Resident trip chain generation method based on multi-source data fusion, and vehicle-sharing query method
CN116431988A (en) * 2023-03-22 2023-07-14 浙江大学 Resident trip activity time sequence generation method based on activity mode-Markov chain

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10649747B2 (en) * 2015-10-07 2020-05-12 Andreas Voellmy Compilation and runtime methods for executing algorithmic packet processing programs on multi-table packet forwarding elements

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108564396A (en) * 2018-03-23 2018-09-21 同济大学 Multimode traffic trip questionnaire survey design method based on D-error design effectivelies
CN110390415A (en) * 2018-04-18 2019-10-29 北京嘀嘀无限科技发展有限公司 A kind of method and system carrying out trip mode recommendation based on user's trip big data
CN111737601A (en) * 2020-06-08 2020-10-02 北京奇虎科技有限公司 Method, device and equipment for recommending travel strategy and storage medium
CN112632374A (en) * 2020-12-18 2021-04-09 东南大学 Resident travel mode selection analysis method considering customized bus
CN113269358A (en) * 2021-05-19 2021-08-17 兆边(上海)科技有限公司 Planning method based on multi-mode integrated travel
WO2023273292A1 (en) * 2021-06-30 2023-01-05 深圳市城市交通规划设计研究中心股份有限公司 Resident trip chain generation method based on multi-source data fusion, and vehicle-sharing query method
CN113780808A (en) * 2021-09-10 2021-12-10 西南交通大学 Vehicle service attribute decision optimization method based on flexible bus connection system line
CN115049217A (en) * 2022-05-17 2022-09-13 中国平安财产保险股份有限公司 Method and device for generating travel strategy, computer equipment and storage medium
CN115412857A (en) * 2022-08-24 2022-11-29 浙江大学 Resident travel information prediction method
CN116431988A (en) * 2023-03-22 2023-07-14 浙江大学 Resident trip activity time sequence generation method based on activity mode-Markov chain

Non-Patent Citations (10)

* Cited by examiner, † Cited by third party
Title
Research on Travel Mode Choice Behavior Under Integrated Multi-modal Transit Information Service;HU Hua，J Teng，YF Gao，XM Zhou;《Transport Policy》;全文 *
Transport decarbonization in big cities: An integrated environmental co-benefit analysis of vehicles purchases quota-limit and new energy vehicles promotion policy in Beijing;Xi Yang，Wanqi Lin，Cecilia Springer;《Sustainable Cities and Society》;全文 *
Yilin Sun ; Zhiyi Huang and Kitamura, R..The Automobility Cohort as a Tool in the Study of Urban Travel.《 2011 International Conference on Electric Technology and Civil Engineering (ICETCE)》.2011,全文. *
基于Two-step Cluste的出行选择行为研究;温惠英;任倩;;《公路工程》(第02期);全文 *
基于两阶段聚类的电动自行车出行者选择研究;胡郁葱;陈杰;邹小健;陈枝伟;;《广西师范大学学报(自然科学版)》(第03期);全文 *
基于聚类分析的铁路出行旅客类别划分;吕红霞;王文宪;蒲松;余大本;;《交通运输系统工程与信息》(第01期);全文 *
家庭收入差异对出行方式选择的影响分析;范琪;王炜;杨洋;华雪东;;《交通信息与安全》(第06期);全文 *
殷复莲.《数据分析与数据挖掘实用教程》.中国传媒大学出版社,2017,第249页. *
活动―出行决策行为与TDM策略互动关系研究的贝叶斯方法;隽志才;宗芳;栾琨;;《交通运输系统工程与信息》(第04期);全文 *
钱兵等.《智能运维之道 基于AI技术的应用实践》.机械工业出版社,2021,第111-113页. *

Also Published As

Publication number Publication date
CN116628527A (en) 2023-08-22

Similar Documents

Publication Publication Date Title
Christoforou et al. Who is using e-scooters and how? Evidence from Paris
Gurumurthy et al. Benefits and costs of ride-sharing in shared automated vehicles across Austin, Texas: Opportunities for congestion pricing
Guidon et al. Electric bicycle-sharing: A new competitor in the urban transportation market? An empirical analysis of transaction data
Circella et al. The adoption of shared mobility in California and its relationship with other components of travel behavior
Reck et al. Explaining shared micromobility usage, competition and mode choice by modelling empirical data from Zurich, Switzerland
Vanoutrive et al. What determines carpooling to workplaces in Belgium: location, organisation, or promotion?
Narayanan et al. Shared mobility services towards Mobility as a Service (MaaS): What, who and when?
Sun et al. Travel behaviours, user characteristics, and social-economic impacts of shared transportation: A comprehensive review
Xiong et al. Understanding operation patterns of urban online ride-hailing services: A case study of Xiamen
Zhou et al. Spatiotemporal characteristics analysis of commuting by shared electric bike: A case study of Ningbo, China
Circella et al. Exploring the relationships among travel multimodality, driving behavior, use of ridehailing and energy consumption
Rafiq et al. An exploratory analysis of alternative travel behaviors of ride-hailing users
Li et al. Factors affecting bike-sharing behaviour in Beijing: price, traffic congestion, and supply chain
Alemi What Makes Travelers Use Ridehailing? Exploring the Latent Constructs behind the Adoption and Frequency of Use of Ridehailing Services, and Their Impacts on the Use of Other Travel Modes
Kim et al. A scenario-based stochastic programming approach for the public charging station location problem
Bi et al. Why they don't choose bus service? Understanding special online car-hailing behavior near bus stops
Shah et al. Why do people take e-scooter trips? Insights on temporal and spatial usage patterns of detailed trip data
CN116628527B (en) Design method and system for integrated travel strategy
Thaithatkul et al. Car versus motorcycle ride-hailing applications: user behaviors and adoption factors in Bangkok, Thailand
Huang et al. To share or not to share? Revealing determinants of individuals’ willingness to share rides through a big data approach
Reck Modelling travel behaviour with shared micro-mobility services and exploring their environmental implications
CN114118874A (en) Intelligent allocation method for electric power trip tasks
Kagho et al. Potential impacts of integrating an on-demand transport service with public transit system: A case study for Zurich
CN112052898A (en) Method and system for constructing potential classification model of intercity high-speed rail passenger
Deneke et al. Transportation Mode Choice Behavior with Multinomial Logit Model: Work and School Trips

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant