Prediction and classification method based on residents' travel influencing factors and transportation resource demand

Abstract

The present invention discloses a prediction and classification method based on factors influencing resident travel and traffic resource demand, comprising the following steps: obtaining multi-source data involved in the urban development process, and constructing a resident travel analysis database under urban development based on the multi-source data; calculating resident travel traffic resource demand based on the database, and constructing a set of factors influencing resident travel resource demand; constructing a resident travel traffic resource demand prediction model based on resident travel traffic resource demand and influencing factors; calculating the importance of resident travel influencing factors based on the resident travel traffic resource demand prediction model, and predicting resident travel resource demand under changes in the influencing factors based on the importance of the resident travel influencing factors, thereby obtaining a prediction and classification result. By accurately predicting resident travel traffic resource demand under changes in multiple factors, the present invention can improve traffic efficiency and promote the sustainable development of urban transportation.

Description

Technical Field

[0001] The present invention belongs to the field of urban traffic travel prediction and classification, and in particular relates to a prediction and classification method based on residents' travel influencing factors and traffic resource demand. Background Art

[0002] With the acceleration of global urbanization, urban populations continue to grow, and cities continue to expand, urban transportation faces increasingly severe challenges. Developing sustainable cities has become a global consensus, and the sustainable development of urban transportation is crucial, impacting residents' quality of life, urban economic activity, and environmental protection. In this context, a deep understanding of the impact of various factors—such as land use, travel scenarios, socioeconomic attributes, and accessibility—on residents' demand for transportation resources in urban development, as well as accurately predicting residents' demand for transportation resources, are crucial for formulating scientific and rational transportation policies, optimizing transportation services, and promoting sustainable urban development.

[0003] During urban development, life events (such as marriage, childbirth, and job changes), the built environment (such as land use, transportation infrastructure, and public transportation accessibility), and contextual factors (such as weather, air pollution, and major public health events) have spatiotemporally and heterogeneously impacted residents' travel behavior. For example, as cities develop, new transportation options such as shared bikes, electric vehicles, and ride-hailing services continue to emerge, changing residents' travel choices. Improved public transportation accessibility and the rise in private car ownership also significantly influence residents' travel patterns. Furthermore, changes in land use, economic growth, evolving social norms, and climate change during urban development indirectly influence residents' work-residence balance, values, and behavioral norms, which in turn impact travel behavior.

[0004] Although a large number of studies have focused on the mechanisms of residents' travel behavior, most of them focus on the impact of a certain type of factor on cross-sectional travel behavior, lacking a comprehensive quantitative analysis of changes in travel behavior over different periods. There is still a lack of a comprehensive framework to systematically analyze the spatiotemporal heterogeneity of the impact of various factors on residents' travel behavior during urban development. There is also insufficient understanding of the changes in the importance of multiple factors on residents' travel behavior during urban development, and the accuracy of forecasting transportation resource demand involving multiple transportation modes during residents' travel is not high. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a prediction and classification method based on residents' travel influencing factors and traffic resource demand, including:

[0006] Acquire multi-source data involved in the urban development process, and construct a resident travel analysis database under urban development based on the multi-source data; the multi-source data includes resident travel data, resident household characteristic data, resident individual characteristic data, urban built environment data, and resident travel scenario data;

[0007] Based on the database, calculating residents' travel transportation resource demands and constructing a set of factors influencing residents' travel transportation resource demands;

[0008] Constructing a prediction model for residents' travel transportation resource demand based on the residents' travel transportation resource demand and influencing factors;

[0009] According to the residents' travel traffic resource demand prediction model, the importance of the residents' travel influencing factors is calculated, and the residents' travel resource demand under the change of the influencing factors is predicted according to the importance of the residents' travel influencing factors to obtain the prediction classification results.

[0010] Preferably, the process of obtaining multi-source data involved in the urban development process and constructing a resident travel analysis database under urban development based on the multi-source data includes:

[0011] Based on two or more sections of multiple time dimensions during the urban development process, sample and collect urban resident travel data representing the travel characteristics of residents in a specific section of urban development, as well as the corresponding resident family characteristics and resident individual characteristics data;

[0012] Determine the temporal and spatial scope of residents' travel based on the collected urban residents' travel data at the corresponding sections, obtain urban built environment data and residents' travel scenario data within the temporal and spatial scope, and combine them to obtain multi-source data;

[0013] Based on multi-source data, a correlation matrix between multi-source data is constructed to obtain a resident travel analysis database covering multi-source heterogeneous data under urban development.

[0014] Preferably, the sampling method includes random sampling, stratified sampling, and address-based sampling methods, and the sampling method is based on which representative sample data of the current section and the current area are obtained.

[0015] Preferably, the temporal and spatial range of residents' travel is the combined range of time and space covered by the urban residents' travel data; the formula is:

[0016]

[0017] Where TSR represents the spatiotemporal range, T represents the time dimension, which is between the earliest travel time T_min and the latest travel time T_max, and S represents the spatial dimension, which belongs to the study area A. A is the convex hull or minimum circumscribed rectangle formed by the starting and ending points of all residents' trips in each section.

[0018] Preferably, the correlation matrix is ​​the correlation relationship between the multi-section resident travel time and the weather data at a specific time in the scene data, and the resident travel location information and the land use data of the corresponding spatial range; the formula expression is:

[0019]

[0020] Where, is the correlation matrix between data sets i and j, is the kth attribute of dataset i, is the lth attribute of dataset j, It is a mapping function between attribute k and attribute l, which is a mapping of temporal relationship or spatial relationship.

[0021] Preferably, the process of calculating residents' travel traffic resource demands based on the database and constructing a set of influencing factors that significantly influence residents' travel traffic resource demands includes:

[0022] Based on the correlation matrix between multi-source data, the full set of factors influencing daily travel resource demand of residents in multiple sections and family characteristics, personal characteristics, scene characteristics, and land use characteristics are extracted;

[0023] According to the complete set of influencing factors, based on the analysis algorithm, the correlation between the influencing factors and the residents' transportation resource demand is calculated;

[0024] A set of factors influencing residents' traffic resource demands is extracted based on the relevance and the preset relevance threshold.

[0025] Preferably, the daily travel resource demand of residents is the distance traveled by residents using various modes of transportation in one day; the formula is:

[0026]

[0027] Where, is the collection of transportation resources required by resident i for one day’s travel, is the distance that resident i travels using transportation mode M in one day.

[0028] Preferably, the formula for extracting the full set of influencing factors is:

[0029]

[0030] Where, It is the extracted feature set of influencing factors, including family characteristics, personal characteristics, scene characteristics, and land use characteristics. represents the kth feature of the jth influencing factor of resident i, is the correlation matrix between resident i’s transportation resource demand and the jth type of influencing factors.

[0031] Preferably, the process of constructing a prediction model for residents' travel traffic resource demand based on the residents' travel traffic resource demand and influencing factors includes:

[0032] According to the residents' transportation resource demand and the set of influencing factors, a deterministic utility function of residents' transportation resource demand under multiple sections is constructed;

[0033] Based on the deterministic utility function, the discrete-continuous transportation resource usage behavior of residents in multiple sections of urban development within a day is simulated, and a transportation resource demand prediction model for residents in multiple sections of urban development is constructed.

[0034] Preferably, the formula expression of the deterministic utility function is:

[0035]

[0036] Where, is the utility of individual i choosing travel mode M, is the household feature vector of individual i, is the personal feature vector of individual i, is the characteristic vector of travel scenarios involving resident i in one day, is the built environment feature vector involved in resident i’s daily trip, 、 、 and is the coefficient vector corresponding to the eigenvector, is the random error term.

[0037] Compared with the prior art, the present invention has the following advantages and technical effects:

[0038] Based on multi-source heterogeneous data such as resident travel data, resident family characteristic data, resident personal characteristic data, urban built environment data, and resident travel scenario data involved in the process of urban development, the present invention constructs a complete database to support the importance analysis of resident travel influencing factors and the prediction of transportation resource demand in the process of urban development; through algorithms such as correlation analysis, a set of influencing factors that have a significant impact on resident transportation resource demand is accurately established; then, a utility function of resident travel resource demand is established to accurately simulate the use behavior of transportation resources covering the discrete-continuous characteristics of residents; finally, a matching mechanism between multi-section resident travel transportation resource demand prediction models is constructed, so as to accurately analyze the changes in the importance of multiple types of influencing factors to resident travel transportation resource demand; according to the variable range of multiple types of influencing factors, a classified prediction of resident travel transportation resource demand under multiple factor changes is realized.

[0039] The present invention can effectively calculate the changes in the importance of multiple factors to residents' travel and transportation resource needs during urban development, accurately predict residents' travel and transportation resource needs under multiple factors, and better assist management departments in formulating transportation supply optimization and demand management policies, improve transportation efficiency, and promote the sustainable development of urban transportation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0041] Figure 1 Schematic diagram of a method flow in an embodiment of the present invention;

[0042] Figure 2 Schematic diagram of the change in absolute value of traffic resource demand under the change of influencing factors in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0044] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0045] like Figure 1 and Figure 2 As shown, this embodiment provides a prediction and classification method based on residents' travel influencing factors and traffic resource demand, including the following steps:

[0046] S1. Acquire multi-source data involved in the urban development process and construct a resident travel analysis database based on the multi-source data; the multi-source data includes resident travel data, resident household characteristic data, resident individual characteristic data, urban built environment data, and resident travel scenario data;

[0047] S2. Based on the database, calculate the residents' travel transportation resource demand and construct a set of factors affecting residents' travel transportation resource demand;

[0048] S3. Construct a prediction model for residents’ travel transportation resource demand based on residents’ travel transportation resource demand and influencing factors;

[0049] S4. Calculate the importance of factors influencing residents' travel based on the residents' travel traffic resource demand prediction model, and predict the residents' travel traffic resource demand under changes in the influencing factors based on the importance of the factors influencing residents' travel to obtain the prediction classification results.

[0050] Furthermore, the process of obtaining multi-source data involved in the urban development process and constructing a resident travel analysis database under urban development based on the multi-source data includes:

[0051] S11. Based on two or more sections across multiple time dimensions during the urban development process, collect sampled urban resident travel data representing the travel characteristics of residents at specific sections of urban development, as well as corresponding data on household and individual characteristics of residents;

[0052] The two end faces or multiple sections of multiple time dimensions can be key periods in the urban development process, such as the start and end years of various plans. This embodiment uses 2014, the year of the fifth travel survey of residents in a certain city, and 2023, the year of the sixth travel survey of residents in a certain city, as the two key sections.

[0053] The urban resident travel data of the resident travel characteristics in a specific section, as well as the corresponding resident family characteristics and resident personal characteristics data are obtained through the resident travel survey data organized by the transportation committee of each city. This embodiment uses the resident travel data, corresponding resident family characteristics and resident personal characteristics data from the fifth and sixth resident travel surveys of a certain city as the data basis.

[0054] S12. Determine the temporal and spatial scope of residents' travel based on the collected urban residents' travel data for the corresponding sections, obtain urban built environment data and residents' travel scenario data within the temporal and spatial scope, and combine them to obtain multi-source data;

[0055] S13. Based on multi-source data, a correlation matrix between multi-source data is constructed to obtain a resident travel analysis database covering multi-source heterogeneous data under urban development.

[0056] Furthermore, the sampling methods include random sampling, stratified sampling, and address-based sampling methods, and the sampling methods are based on sampling to obtain representative sample data including the current section and the current area.

[0057] Furthermore, the spatiotemporal range of residents' travel is the combined range of time and space covered by the urban residents' travel data; the formula is:

[0058]

[0059] Where TSR represents the spatiotemporal range, T represents the time dimension, which is between the earliest travel time T_min and the latest travel time T_max, and S represents the spatial dimension, which belongs to the study area A. A is the convex hull or minimum circumscribed rectangle formed by the starting and ending points of all residents' trips in each section.

[0060] Multi-source data such as urban built environment data and residents' travel scenario data within the time and space scope can be obtained through public data such as Amap crawling, weather network, and air quality monitoring station public data.

[0061] Furthermore, the correlation matrix is ​​the correlation between the multi-section resident travel time and the weather data at a specific time in the scene data, and the resident travel location information and the land use data of the corresponding spatial range; the formula expression is:

[0062]

[0063] Where, is the correlation matrix between data sets i and j, is the kth attribute of dataset i, is the lth attribute of dataset j, It is a mapping function between attribute k and attribute l, which is a mapping of temporal relationship or spatial relationship.

[0064] Furthermore, based on the database, the process of calculating residents' travel transportation resource demands and constructing a set of factors that significantly affect residents' travel transportation resource demands includes:

[0065] S21. Based on the correlation matrix between multi-source data, extract the complete set of factors affecting residents' daily travel transportation resource demand and family characteristics, personal characteristics, scene characteristics, and land use characteristics in multiple sections;

[0066] S22. Based on the complete set of influencing factors and the analysis algorithm, calculate the correlation between the influencing factors and residents' transportation resource demands;

[0067] Correlation analysis algorithms include methods such as the Pearson correlation coefficient and the Spiegel correlation coefficient to measure the correlation between continuous influencing factors and residents' travel transportation resource demand. They can also be methods such as Cohen's d and KL divergence to analyze the differences in residents' travel transportation resource demand under multiple influencing factors.

[0068] S23. Extract a set of factors influencing residents' traffic resource demands based on the relevance and the preset relevance threshold.

[0069] The preset correlation threshold can be a correlation threshold such as ±0.6 to measure whether there is a large correlation, or it can be a significance index such as t-test, P value to measure whether there is a significant difference in residents' travel transportation resource demand under multiple influencing factors.

[0070] Furthermore, the daily travel resource demand of residents is the distance they travel using various modes of transportation in a day; the formula is:

[0071]

[0072] Where, is the collection of transportation resources required by resident i for one day’s travel, is the distance that resident i travels using transportation mode M in one day.

[0073] Furthermore, the formula for extracting the full set of influencing factors is:

[0074]

[0075] Where, It is the extracted feature set of influencing factors, including family characteristics, personal characteristics, scene characteristics, and land use characteristics. represents the kth feature of the jth influencing factor of resident i, is the correlation matrix between resident i’s transportation resource demand and the jth type of influencing factors.

[0076] Furthermore, based on residents' travel transportation resource demand and influencing factors, the process of constructing a residents' travel transportation resource demand prediction model includes:

[0077] S31. Based on the residents' transportation resource demand and the set of influencing factors, a deterministic utility function of residents' transportation resource demand under multiple sections is constructed;

[0078] S32. Based on the deterministic utility function, simulate the discrete-continuous transportation resource usage behavior of residents in multiple sections of urban development within a day, and construct a transportation resource demand prediction model for residents in multiple sections of urban development.

[0079] Furthermore, the formula of the deterministic utility function is:

[0080]

[0081] Where, is the utility of individual i choosing travel mode M, is the household feature vector of individual i, is the personal feature vector of individual i, is the characteristic vector of travel scenarios involving resident i in one day, is the built environment feature vector involved in resident i’s daily trip, 、 、 and is the coefficient vector corresponding to the eigenvector, is the random error term.

[0082] The daily discrete-continuous transportation resource usage behavior of residents across multiple sections can be simulated using a variety of methods, including multiple discrete-continuous extreme value models, two-stage decision models, Markov chain models, and game theory models. This embodiment uses a multiple discrete-continuous extreme value model to simulate the daily discrete-continuous transportation resource usage behavior of residents across multiple sections.

[0083] The multiple discrete continuous extreme value model aims to maximize the utility of resident i's use of transportation resources within a day. The utility formula is:

[0084]

[0085] Where, It is a quasi-concave, monotonically increasing and continuously differentiable function of the resource vectors used by various transportation modes; is the resource usage of transportation mode m, in this embodiment, the unit is km; and is the saturation parameter of mode m, the difference is that, The important role of is to adjust the indifference curve and make the corner solution (that is, the resource consumption of mode m is zero) possible. The intuitive explanation is that The higher the value of , the stronger the residents' tendency to use mode m. The difference is that Saturation is controlled by indexing the resource consumption of mode m, that is, as the resource consumption of mode m increases, the marginal utility decreases; represents the baseline marginal utility of mode m, in other words, the marginal utility at the point of zero resource consumption of mode m. Usually, use in the form of It is the determining factor of the basic utility of each method. are independent and identically distributed error terms that follow the Gambull distribution. In addition, is a function that captures the influence of the characteristics of the decision maker and the alternatives, including a constant to capture the general preference for option m, and is the coefficient to be estimated.

[0086] To simulate the discrete-continuous transportation resource usage behavior of residents within a day under multiple sections of urban development, the multiple discrete-continuous extreme value model adopted in this embodiment needs to maximize the utility of resident i using various transportation resources, which is expressed as follows:

[0087]

[0088] Where, is the unit quantity of the usage mode m, which is km in this embodiment; is the total amount of resources used by resident i in a day by each mode of transportation. In this example, it is the total distance resident i travels in a day. To obtain the optimal transportation resource allocation for resident i, the above optimization problem can be solved by constructing a Lagrangian function and applying the Kuhn-Tucker condition, further generating the following expressions for the resource consumption probabilities of each mode of transportation.

[0089]

[0090] Where N is the number of transportation modes consumed;

[0091] , .

[0092] According to the probability expression of resource consumption of resident i using each transportation mode, this embodiment performs parameter estimation through maximum likelihood estimation, Bayesian estimation and other methods. In this embodiment, maximum likelihood estimation is used for estimation, and the parameter estimation results are shown in Table 1:

[0093] Table 1

[0094] variable Impact 2014 Coefficient estimation results (t-values ​​in brackets) Impact in 2023 Coefficient estimation results (t-values ​​in brackets) Gender: Male private car 0.721（29.994） private car 0.742（22.733） Age: 26-50 private car 0.463(16.988) private car 0.900(22.030) Annual family income: less than 100,000 yuan walk 0.069(3.374) walk 0.222(10.149) Family with bicycles Ground public transportation 0.024(1.169) Ground public transportation -0.073(-2.046) Living in the suburbs subway -0.074(-1.848) subway 0.274(4.795) Full-day subway travel time (min) subway -0.002(-19.822) subway -0.001(-11.374) Severe air pollution private car 0.085(2.423) private car 0.960(3.136) Rainy Day Ground public transportation -0.057(-2.557) Ground public transportation 0.105(2.725)

[0095] Furthermore, based on the residents' travel traffic resource demand prediction model, the importance of the factors affecting residents' travel is calculated, and based on the importance of the factors affecting residents' travel, the residents' traffic resource demand under the change of the factors is predicted. The process of obtaining the prediction classification results includes:

[0096] S41. Based on the multi-section resident travel resource demand prediction model under urban development, a matching mechanism for the prediction models among multiple sections is designed, and the importance of influencing factors among comparable multiple sections of urban development is calculated.

[0097] The matching mechanism of the multi-section resident transportation resource demand prediction model in this embodiment is Markov state transition estimation, variable value standardization based on GDP growth rate, and marginal substitution rate calculation based on time equivalence. In this embodiment, it is assumed that one minute of residents' time between 2014 and 2023 is equally important. The full-day subway travel time (min) is used as the matching mechanism for multi-section resident transportation resource demand. The marginal substitution rate of multiple influencing factors is calculated as the importance of influencing factors among multiple sections of urban development. Since personal and family economic attribute variables are not within the controllable scope of urban transportation management departments, the calculation results of the importance of household vehicle ownership, built environment, and scenario variables are shown in Table 2:

[0098] Table 2

[0099] variable Impact 2014 Coefficient estimation results (t-values ​​in brackets) Impact in 2023 Coefficient estimation results (t-values ​​in brackets) Family with bicycles Ground public transportation -12.073 (-1.190) Ground public transportation 69.159 (2.003) Living in the suburbs subway 291.777 (15.888) subway -11.112 (-0.314) Full-day subway travel time (min) subway 1 (baseline value) subway 1 (baseline value) Severe air pollution private car -42.465 (-2.365) private car -915.99 (-2.947) Rainy Day Ground public transportation 28.132 (2.585) Ground public transportation -100.46 (-2.649)

[0100] S42. Based on the multi-section resident travel traffic resource demand prediction model under urban development, simulate the possible changes in the factors affecting resident travel traffic resource demand, predict the resident traffic resource demand under the changes in the influencing factors, and classify them to obtain the prediction and classification results.

[0101] Possible changes in factors affecting residents' travel and transportation resource demand can be caused by the redistribution of residents' residences and changes in the balance between work and residence due to land use layout, or by the improvement of road network density and public transportation accessibility due to the optimization of urban transportation infrastructure. This example takes the improvement of all-day subway travel time as an example, and calculates the multiple change and absolute value change of the number of people using the subway in a certain area of ​​the city, as well as the transportation resource demand, respectively. Figure 2 shown.

[0102] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A prediction and classification method based on factors affecting residents' travel and traffic resource demand, characterized by: include: Acquire multi-source data involved in the urban development process, and construct a resident travel analysis database under urban development based on the multi-source data; the multi-source data includes resident travel data, resident household characteristic data, resident individual characteristic data, urban built environment data, and resident travel scenario data; Based on the database, calculating residents' travel transportation resource demands and constructing a set of factors influencing residents' travel transportation resource demands; Constructing a prediction model for residents' travel transportation resource demand based on the residents' travel transportation resource demand and influencing factors; Calculating the importance of factors influencing residents' travel based on the residents' travel traffic resource demand prediction model, and predicting residents' traffic resource demand under changes in the influencing factors based on the importance of the factors influencing residents' travel to obtain a prediction classification result; The process of obtaining multi-source data involved in the urban development process and constructing a resident travel analysis database under urban development based on the multi-source data includes: Based on two or more sections of multiple time dimensions during the urban development process, sample and collect urban resident travel data representing the travel characteristics of residents in a specific section of urban development, as well as the corresponding resident family characteristics and resident individual characteristics data; Determine the temporal and spatial scope of residents' travel based on the collected urban residents' travel data at the corresponding sections, obtain urban built environment data and residents' travel scenario data within the temporal and spatial scope, and combine them to obtain multi-source data; Based on multi-source data, a correlation matrix between multi-source data is constructed to obtain a resident travel analysis database covering multi-source heterogeneous data under urban development; The temporal and spatial scope of residents' travel is the combined time and space scope covered by the urban residents' travel data; the formula is: Where TSR represents the spatiotemporal range, T represents the time dimension, which is between the earliest travel time T_min and the latest travel time T_max, and S represents the spatial dimension, which belongs to the study area A. A is the convex hull or minimum circumscribed rectangle formed by the starting and ending points of all residents' trips in each section.

2. The method according to claim 1, characterized in that The sampling methods include random sampling, stratified sampling, and address-based sampling methods. Based on the sampling methods, representative sample data of the current section and the current area are obtained.

3. The method according to claim 1, characterized in that The correlation matrix is ​​the correlation between the multi-section resident travel time and weather data in the scene data, the resident travel location information and the land use data of the corresponding spatial range; the formula expression is: Where, is the correlation matrix between data sets i and j, is the kth attribute of dataset i, is the lth attribute of dataset j, It is a mapping function between attribute k and attribute l, which is a mapping of temporal relationship or spatial relationship.

4. The method according to claim 1, wherein The process of calculating residents' travel transportation resource demands based on the database and constructing a set of influencing factors that significantly influence residents' travel transportation resource demands includes: Based on the correlation matrix between multi-source data, the full set of factors influencing daily travel resource demand of residents in multiple sections and family characteristics, personal characteristics, scene characteristics, and land use characteristics are extracted; According to the complete set of influencing factors, based on the analysis algorithm, the correlation between the influencing factors and the residents' transportation resource demand is calculated; A set of factors influencing residents' traffic resource demands is extracted based on the relevance and the preset relevance threshold.

5. The method according to claim 4, characterized in that The daily travel resource demand of residents is the distance they travel using various modes of transportation in one day; the formula is: Where, is the collection of transportation resources required by resident i for one day’s travel, is the distance that resident i travels using transportation mode M in one day.

6. The method according to claim 4, characterized in that The formula for extracting the full set of influencing factors is: Where, It is the extracted feature set of influencing factors, including family characteristics, personal characteristics, scene characteristics, and land use characteristics. represents the kth feature of the jth influencing factor of resident i, is the correlation matrix between resident i’s transportation resource demand and the jth type of influencing factors.

7. The method according to claim 1, characterized in that Based on the above-mentioned residents' travel transportation resource demand and influencing factors, the process of constructing a residents' travel transportation resource demand prediction model includes: According to the residents' transportation resource demand and the set of influencing factors, a deterministic utility function of residents' transportation resource demand under multiple sections is constructed; Based on the deterministic utility function, the discrete-continuous transportation resource usage behavior of residents in multiple sections of urban development within a day is simulated, and a transportation resource demand prediction model for residents in multiple sections of urban development is constructed.

8. The method according to claim 7, characterized in that The formula expression of the deterministic utility function is: Where, is the utility of individual i choosing travel mode M, is the household feature vector of individual i, is the personal feature vector of individual i, is the characteristic vector of travel scenarios involving resident i in one day, is the built environment feature vector involved in resident i’s daily trip, 、 、 and is the coefficient vector corresponding to the eigenvector, is the random error term.

Images