Urban built environment and climate factor and network car demand correlation evaluation system

Abstract

The present application relates to the technical field of traffic demand management, and discloses a system for evaluating the correlation between urban built environment and climate factors and online car-hailing demand, comprising: an input module for obtaining grid-based urban built environment and climate data and corresponding online car-hailing demand data, and performing data cleaning and screening; an index module for calculating and storing a built environment index system related to online car-hailing demand, and quantitatively calculating the index values in the built environment index system according to the data input by the input module; an analysis module for establishing a multi-layer development model for analyzing the correlation between urban built environment and climate factors and online car-hailing demand according to the index values output by the index module; and an evaluation output module for obtaining evaluation results and performing analysis through the multi-layer development model. The present application comprehensively evaluates the correlation between various factors and online car-hailing travel demand, optimizes online car-hailing scheduling and resource allocation during urban construction, and improves the efficiency of online car-hailing operation services.

Description

Technical Field

[0001] This invention relates to the field of traffic demand management technology, specifically to an assessment system for the correlation between urban built environment and climate factors and ride-hailing demand. Background Technology

[0002] Ride-hailing services such as Didi Chuxing have developed rapidly in recent years as an important mode of shared mobility, providing people with a more convenient and efficient travel experience.

[0003] Currently, most studies on the correlation between ride-hailing demand and various environmental factors employ statistical methods. Schaller, B. (2005) found that subway commuting demand, the number of households without cars, and the number of taxi orders originating from or terminating at airports are the three most important factors predicting urban taxi demand. Gonzales, E. (2014) found that public transport accessibility, population size, and income are the main explanations for taxi passenger volume. Yang, Z. (2018) et al., through studying the impact of land use and accessibility of various travel modes on taxi demand, found that there is no significant positive correlation between mixed land use and taxi demand, and that taxis and subways have a significant complementary relationship, while taxis compete with public transport. Lacombe, A. (2014) found that income, age, time required to reach public transport stations, and the number of parking spaces in the area are the main factors influencing taxi demand.

[0004] Current research mostly analyzes the correlation between different factors and taxi travel demand from a limited perspective, failing to comprehensively consider the combined effects of different types, especially time-varying and time-invariant factors; it also fails to optimize taxi or ride-hailing dispatch and resource allocation in the process of urban construction, resulting in low efficiency of ride-hailing operation services.

[0005] Therefore, establishing a unified model that encompasses multiple factors and conducts a comprehensive evaluation has significant theoretical and practical value. Summary of the Invention

[0006] To address the problems existing in the prior art, the purpose of this invention is to provide an assessment system for the correlation between urban built environment and climate factors and ride-hailing demand, which comprehensively evaluates the correlation between multiple factors and ride-hailing travel demand; optimizes ride-hailing scheduling and resource allocation in the process of urban construction, and improves the efficiency of ride-hailing operation services.

[0007] To achieve the above objectives, the present invention employs the following technical solutions.

[0008] A system for assessing the correlation between urban built environment and climate factors and ride-hailing demand includes:

[0009] The input module is used to acquire gridded urban built environment and climate data, as well as corresponding ride-hailing demand data, and to perform data cleaning and filtering; and then input the cleaned and filtered data into the indicator module.

[0010] The indicator module is used to calculate and store the built environment indicator system related to ride-hailing demand, and to quantify the indicator values ​​of multiple indicators in the built environment indicator system based on the data input from the input module.

[0011] The analysis module establishes a multi-layered development model to analyze the correlation between urban built environment and climate factors and ride-hailing demand, based on the indicator values ​​output by the indicator module.

[0012] The assessment output module is used to obtain assessment results on the correlation between urban built environment and climate factors and ride-hailing demand through a multi-layer development model, and to analyze the results.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0014] (1) The model of this invention is compatible with both time-varying and time-invariant variables, and can explore the interaction between different variables, making it more systematic.

[0015] (2) The present invention has high flexibility and can not only handle the problem of different number of repeated measurements of the research objects, but also handle the problem of different time intervals between repeated measurements of each object, making it more user-friendly in terms of data acquisition and cleaning.

[0016] (3) This invention can detect and reflect the demand change trend over a long period of time, help verify the process of the spatiotemporal change of the correlation between each explanatory variable and the dependent variable, and improve the scientific nature of the analysis. Attached Figure Description

[0017] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0018] Figure 1 This is a flowchart of the process of the present invention;

[0019] Figure 2 A map showing the distribution of Didi Express passenger pick-up points in Shanghai;

[0020] Figure 3 This is a map showing the distribution of taxi pick-up points in Shanghai.

[0021] Figure 4 It is a map of Shanghai's urban area that combines a 1km×1km grid with urban expressways. Detailed Implementation

[0022] The embodiments of the present invention will be described in detail below with reference to examples. However, those skilled in the art will understand that the following examples are only for illustrating the present invention and should not be regarded as limiting the scope of the present invention.

[0023] This embodiment uses Shanghai (excluding Chongming District) as the analysis object. The data source is Didi Express / Taxi order data from November 30, 2016 to February 28, 2017, with data during the Spring Festival period being excluded as atypical data.

[0024] refer to Figure 1 This is a flowchart of the workflow of the present invention. A system for assessing the correlation between urban built environment and climate factors and ride-hailing demand includes:

[0025] The input module is used to acquire gridded urban built environment and climate data, as well as corresponding ride-hailing demand data, and to perform data cleaning and filtering; and then input the cleaned and filtered data into the indicator module.

[0026] Data acquisition in the input module is achieved through the following steps:

[0027] Step 1.1: Obtain urban built environment data and grid the city using a 1km×1km grid; combine the urban expressways to divide the city into the core urban area, the middle urban area, the outer urban area, and the suburban area.

[0028] Step 1.2: Obtain information on different land use types, public transportation station distribution, and routes included in public transportation stations within each grid using built environment data; the public transportation stations include bus stops and subway stations.

[0029] Step 1.3: Obtain climate data for each grid through urban weather stations. The climate data includes temperature and rainfall data for different times of the day.

[0030] Step 1.4: Count the number of ride-hailing order origin points in each grid.

[0031] The order data was cleaned and filtered. The explanatory variables related to ride-hailing demand selected in the input module are:

[0032] 1. The week number corresponding to the measurement date; where the measurement date is Wednesday of each week, and data is taken from Wednesday of each week;

[0033] 2. Weather; whether it will rain;

[0034] 3. Temperature; including daytime and nighttime temperatures;

[0035] 4. The city tier where the grid is located;

[0036] 5. Accessibility to subway and bus networks;

[0037] 6. Land use patterns of different types within the grid, including commercial land, residential land, transportation land, industrial land, educational land, government land, and green space.

[0038] The indicator module is used to calculate and store the built environment indicator system related to ride-hailing demand, and to quantify the indicator values ​​of multiple indicators in the built environment indicator system based on the data input from the input module.

[0039] The cleaned and filtered data input into the input module is quantified to facilitate subsequent model calculations. The cleaned and filtered data specifically includes two major categories of indicators: time-varying variables and time-invariant variables, as shown in Table 1.

[0040] Table 1 Explanatory Variables for the Indicator Module

[0041]

[0042] The calculation of indicator values ​​for multiple indicators in the built environment indicator system within the indicator module is achieved through the following steps:

[0043] Step 2.1: Perform spatial analysis using ArcGIS software to calculate the percentage of land area of ​​different land use types within each grid, and obtain the land use vector.

[0044] Step 2.2: Calculate the bus stop accessibility for each grid to obtain the bus stop accessibility index for each grid.

[0045] The specific method for calculating the bus stop accessibility for each grid is as follows:

[0046] Step 2.21: Calculate the percentage of the total grid area that is within a 500m radius of a single bus stop.

[0047] Step 2.22: Multiply the percentage by the number of bus routes included in the bus stop to obtain the contribution value of a single bus stop to the accessibility of the grid bus stops;

[0048] Step 2.23: The contribution values ​​of all bus stops within a grid to that grid are summed to obtain the bus stop accessibility index of the grid, namely the bus service index (BSI) within the grid.

[0049] The Bus Service Index (BSI) for public transport service levels within a grid is calculated as follows:

[0050]

[0051] Among them, BSI i x represents the public transport service index within the i-th grid;ij It is the percentage of the area within a 500m radius of bus stop j that is located within grid i; α j This is the number of bus routes that pass through station j.

[0052] Step 2.3: Obtain subway station information within each grid.

[0053] The analysis module establishes a multi-layered development model to analyze the correlation between urban built environment and climate factors and ride-hailing demand, based on the indicator values ​​output by the indicator module.

[0054] The establishment of the multi-layered development model in the analysis module is achieved through the following steps:

[0055] Step 3.1: Combine the climate data and urban built environment data in the input module, as well as the ride-hailing data and the built environment indicators in the indicator module to obtain a nested dataset that integrates multiple explanatory variable indicators and time series data.

[0056] Step 3.2: Perform pre-analysis on the data in the nested dataset, and predict the time-varying and time-invariant variables in the explanatory variables, as well as the interaction between the two types of variables;

[0057] Time-varying variables include week number, weather, daytime temperature, and nighttime temperature;

[0058] Time-invariant variables include urban clusters, subway and public transport accessibility, and different types of land use.

[0059] Step 3.3: Build a test model using SAS, determine the random effects in the test model using multiple information evaluation indicators (AIC, BIC, -2LL), and compare the estimation methods (REML or ML).

[0060] Step 3.4 involves progressively improving the test model, adjusting the explanatory variables and interaction terms covered by the test model, and obtaining the final multi-layered development model.

[0061] The two-level development model used treats the divided grids as independent observation objects. Level 1 units are repeated observations within each grid over time, including demand and weather changes at each time point. Level 2 units are the grids themselves, whose attributes include indicators such as the built environment and accessibility of the grid's location. The intra-variable and inter-variable model representations constituting the two-level model are as follows:

[0062] Level 1:

[0063] y ij =π 0j +π 1j t ij +ε ij ,

[0064] Level 2:

[0065]

[0066]

[0067] Where y ij Let t be the observation value in grid j at time node i. ij The corresponding time point; the intercept coefficient π in the model 0j and slope coefficient π 1j All include random effects that follow a bivariate normal distribution; ε ij Represents the observed value y ij random effects, ε ij Follows a bivariate normal distribution It is a bivariate normal distribution The parameters; u 0j The intercept coefficient π 0j random effects; u 1j The slope coefficient π 1j random effects; u 0j and u 1j Follows a bivariate normal distribution T β It is a bivariate normal distribution The parameter; x j Let β be the explanatory variable in grid j. 00 For x j The mean intercept, β 01 β is the regression slope; 10 For x j The average slope, β 11 The slope is the regression slope.

[0068] The assessment output module is used to obtain assessment results on the correlation between urban built environment and climate factors and ride-hailing demand through a multi-layer development model, and to analyze the results.

[0069] The evaluation output module obtains and analyzes evaluation results, including the following steps:

[0070] Step 4.1: Load data into the multi-layered development model and obtain the evaluation results;

[0071] Step 4.2: Analyze the explanatory effects of different levels and types of explanatory variables on within-group and between-group variation in the multilevel development model. Specific implementation examples:

[0073] Figure 2 This is a map showing the distribution of Didi Express passenger pick-up points in Shanghai. Figure 3 This is a map showing the distribution of taxi pick-up points in Shanghai. Figure 2 and Figure 3 It can be seen that the density of Didi Express orders is significantly higher than that of taxis. However, in suburban areas (Figure 2 and...), the density is lower. Figure 3 The areas marked 1 and 2 show a significantly higher density of taxi orders. This may be due to Shanghai's restrictions on taxis with Shanghai C license plates entering the inner ring road.

[0074] Figure 4 This is a map of Shanghai's urban area combining a 1km×1km grid and urban expressways. Grids located on expressways and some grids with low ride-hailing demand (mainly in suburban areas) were removed. The number of grids in the core area, middle area, outer area, and suburbs are 86, 125, 239, and 947, respectively. A total of 316,371 orders were selected over 11 sample days.

[0075] In this example, the time series data of ride-hailing order volume calculated based on urban grids is used as the outcome measure. The analysis uses the MIXED module in SAS software. The model adopts a RIS-AR(1) structure, in which the intercept and slope of the grid level both have random terms. At the same time, AR(1), as a commonly used and effective structure in processing time series data, is used for the covariance structure within the multi-day data in each grid.

[0076] The estimation results obtained by combining SAS analysis are shown in Tables 2, 3 and 4. In this example, three models were gradually established based on the types of explanatory variables: Model A, Model B and Model C.

[0077] Model A included random terms for both the intercept and slope, primarily to examine the influence of concentric circle factors and the interaction effects between time and concentric circles. Model B added time-invariant variables to Model A, including indicators measuring public transportation coverage and land use indicators within the grid. Finally, Model C further incorporated climate factors. For each model, -2LL (-2log likelihood), AIC (Akaike Information Criterion), and BIC (Bayesian Information Criterion) indices were obtained to verify the model improvements. The fixed effects estimation results for the model variables of Didi Express, taxis, and the total are shown in Tables 2, 3, and 4, respectively.

[0078] Table 2. Fixed effects estimation results of model variables for Didi Express.

[0079]

[0080] Note: ~p≤0.10; *p≤0.05; **p≤0.01; ***p≤0.001.

[0081] TA: Public transport accessibility; LU: Land use; W: Weather

[0082] Table 3. Fixed effects estimation results for taxi model variables.

[0083]

[0084] Note: ~p≤0.10; *p≤0.05; **p≤0.01; ***p≤0.001.

[0085] TA: Public transport accessibility; LU: Land use; W: Weather

[0086] Table 4 shows the fixed effects estimation results for the total model variables.

[0087]

[0088] Note: ~p≤0.10; *p≤0.05; **p≤0.01; ***p≤0.001.

[0089] TA: Public transport accessibility; LU: Land use; W: Weather

[0090] Based on the fixed-effects estimation results, it can be seen that within the time frame covered by this example, Didi Express's business volume declined significantly, while taxi business volume increased accordingly. This may be due to Shanghai's implementation of the "Shanghai-registered, Shanghai-vehicle" management measure for ride-hailing services during this period. Furthermore, the distribution of Didi Express orders exhibits a significant stratification effect. In terms of public transportation, the subway and Didi Express demonstrate greater complementarity. Regarding land use, Didi Express demand shows a significant positive correlation with commercial and residential land use, while the correlation with industrial land use is negative. Rainfall also shows a significant positive correlation with Didi Express demand, while daytime and nighttime temperatures show positive and negative correlations, respectively. This invention comprehensively assesses the correlations between multiple factors and ride-hailing travel demand; optimizes ride-hailing dispatching and resource allocation during urban construction, thereby improving the efficiency of ride-hailing operations.

[0091] Although the present invention has been described in detail in this specification with general description and specific embodiments, some modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention are within the scope of protection claimed by the present invention.

Claims

1. A system for assessing the correlation between urban built environment and climate factors and ride-hailing demand, characterized in that, include: The input module is used to acquire gridded urban built environment and climate data, as well as corresponding ride-hailing demand data, and to perform data cleaning and filtering; and then input the cleaned and filtered data into the indicator module. The data acquisition in the input module is achieved through the following steps: Step 1.1: Obtain urban built environment data and grid the city using a 1km×1km grid; combine the urban expressways to divide the city into the core urban area, the middle urban area, the outer urban area, and the suburban area. Step 1.2: Obtain information on different land use types, public transportation station distribution, and routes included in public transportation stations within each grid using built environment data; the public transportation stations include bus stops and subway stations. Step 1.3: Obtain climate data for each grid through urban weather stations. The climate data includes temperature and rainfall data for different times of the day. Step 1.4: Count the number of ride-hailing order origin points in each grid. The indicator module is used to calculate and store the built environment indicator system related to ride-hailing demand, and to quantify the indicator values ​​of multiple indicators in the built environment indicator system based on the data input from the input module. The calculation of the index values ​​of multiple indicators in the built environment index system in the index module is achieved through the following steps: Step 2.1: Perform spatial analysis using ArcGIS software to calculate the percentage of land area of ​​different land use types within each grid, and obtain the land use vector. Step 2.2: Calculate the bus stop accessibility for each grid to obtain the bus stop accessibility index for each grid. The specific method for calculating the bus stop accessibility of each grid is as follows: Step 2.21: Calculate the percentage of the total grid area that is within a 500m radius of a single bus stop. Step 2.22: Multiply the percentage by the number of bus routes included in the bus stop to obtain the contribution value of a single bus stop to the accessibility of the grid bus stops; Step 2.23: The contribution values ​​of all bus stops within the grid to the grid are summed to obtain the bus stop accessibility index of the grid, namely the bus service index (BSI) within the grid. Step 2.3: Obtain subway station information within each grid. The analysis module establishes a multi-layered development model to analyze the correlation between urban built environment and climate factors and ride-hailing demand, based on the indicator values ​​output by the indicator module. The assessment output module is used to obtain assessment results on the correlation between urban built environment and climate factors and ride-hailing demand through a multi-layer development model, and to analyze the results.

2. The system for assessing the correlation between urban built environment and climate factors and ride-hailing demand according to claim 1, characterized in that, The cleaned and filtered data includes the following explanatory variables: the week number corresponding to the measurement day, weather, temperature, the city zoning of the grid, subway and public transport accessibility of the grid, and different types of land use in the grid.

3. The system for assessing the correlation between urban built environment and climate factors and ride-hailing demand according to claim 1, characterized in that, The establishment of the multi-layered development model in the analysis module is achieved through the following steps: Step 3.1: Combine the climate data and urban built environment data in the input module, as well as the ride-hailing data and the built environment indicators in the indicator module to obtain a nested dataset that integrates multiple explanatory variable indicators and time series data. Step 3.2: Perform pre-analysis on the data in the nested dataset, and predict the time-varying and time-invariant variables in the explanatory variables, as well as the interaction between the two types of variables; Step 3.3: Build a test model using SAS, and use various information evaluation indicators such as AIC, BIC, and -2LL to determine the random effects in the test model, and compare the estimation methods REML or ML. Step 3.4 involves progressively improving the test model, adjusting the explanatory variables and interaction terms covered by the test model, and obtaining the final multi-layered development model.

4. The system for assessing the correlation between urban built environment and climate factors and ride-hailing demand according to claim 3, characterized in that, In step 3.2: Time-varying variables include week number, weather, daytime temperature, and nighttime temperature; Time-invariant variables include urban clusters, subway and public transport accessibility, and different types of land use.

5. The system for assessing the correlation between urban built environment and climate factors and ride-hailing demand according to claim 3, characterized in that, In step 3.4, a two-level, two-layer development model is used, with each layer of the grid as an independent observation object. The intra-variable and inter-variable models of the two-layer development model are as follows: Level 1: ， Level 2: , in For the observation in grid j at time node i, For the corresponding time points; the intercept coefficients in the model. and slope coefficient All of them include random effects that follow a bivariate normal distribution; Represents the observed value random effects Follows a bivariate normal distribution ; It is a bivariate normal distribution Parameters; Intercept coefficient random effects; Slope coefficient random effects; and Follows a bivariate normal distribution , It is a bivariate normal distribution Parameters; For the explanatory variables in grid j, for The average intercept, The regression slope; for The average slope, The slope is the regression slope.

6. The system for assessing the correlation between urban built environment and climate factors and ride-hailing demand according to claim 1, characterized in that, The evaluation output module obtains and analyzes evaluation results, including the following steps: Step 4.1: Load data into the multi-layered development model and obtain the evaluation results; Step 4.2: Analyze the explanatory effects of different levels and types of explanatory variables on within-group and between-group variation in the multilevel development model.

Images