A station-city integration maturity evaluation method and system

Abstract

The application provides a station-city integration maturity evaluation method and system, comprising the following steps: constructing a standardized multi-dimensional evaluation index system based on multi-source heterogeneous city data, combining an entropy weight method and a feature importance to realize objective weight calibration, introducing three types of deep learning models to extract spatial, time sequence and semantic implicit features, calculating maturity scores through "index weighting + deep feature" dual-channel fusion, and matching a dynamic updating mechanism and a full-process automatic system to realize objective quantification, scalability and dynamic updating of intelligent evaluation of station-city integration maturity.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation planning technology, and in particular to a method and system for evaluating the maturity of station-city integration. Background Technology

[0002] Currently, the evaluation of station-city integration mostly adopts subjective scoring by experts, static indicator calculation, or planning experience judgment, and has not yet formed a standardized, intelligent, and dynamically updated mature evaluation system. The existing technology has the following core technical defects: 1. There is a lack of a unified and standardized evaluation framework, and a lack of systematic integration mechanisms for multi-source heterogeneous data. The existing evaluation indicators for station-city integration lack a unified mathematical definition in terms of dimensions and calculation methods. Different cities and different routes use different evaluation systems, making it impossible to form a standardized maturity evaluation framework that can be compared horizontally. At the same time, the evaluation data sources for passenger flow, spatial structure, functional formats, and population heat are scattered and highly heterogeneous, lacking a systematic data fusion and processing model, making it difficult to support multi-dimensional unified quantitative evaluation.

[0003] 2. The setting of indicator weights is highly subjective and lacks a data-driven objective optimization and calibration mechanism. The weights of existing evaluation methods often rely on manual setting or determination by the analytic hierarchy process, which are significantly influenced by subjective experience. They lack objective calculation and algorithm optimization mechanisms based on the characteristics of the data itself, which can easily lead to biases in the evaluation results and insufficient overall stability and reproducibility.

[0004] 3. Insufficient ability to characterize complex spatiotemporal coupling features, and lack of intelligent nonlinear evaluation and modeling methods. Station-city integration is a complex urban system involving the coupling of multiple elements such as space, transportation, function, and passenger flow. Existing static evaluation methods cannot effectively characterize the nonlinear implicit features such as the spatial topological relationship of stations, the temporal dynamics of passenger flow, and the long-range dependence between stations. They also lack the ability to automatically infer based on intelligent models and are difficult to adapt to the multi-indicator coupled calculation needs under complex urban systems.

[0005] 4. The evaluation model is static and cannot match the dynamic evolution of station-city integration. Existing evaluations are mostly single static calculations, lacking the ability to dynamically depict the evolution of the station-city integration level over time. They cannot use spatiotemporal sequence data to capture the dynamic changes in the functions and vitality around the station, and the evaluation results are difficult to match the real-time development status of urban stations.

[0006] 5. Lack of a fully automated evaluation system results in insufficient evaluation efficiency and reusability. Existing systems mostly remain at the level of static data display, and have not formed a fully automated processing chain from data collection, indicator calculation, model training to result output. The evaluation process involves a lot of manual intervention, has low execution efficiency, and poor scenario reusability, and cannot support the large-scale evaluation needs of a large number of sites. Summary of the Invention

[0007] This invention provides a method and system for evaluating the maturity of station-city integration, in order to address the deficiencies in existing technologies.

[0008] In a first aspect, the present invention provides a method for evaluating the maturity of station-city integration, comprising: Acquire multi-source heterogeneous data of rail transit stations and their surrounding areas and perform standardized preprocessing. The multi-source heterogeneous data includes rail transit passenger flow data, land use data, points of interest data, road network data, and population density data. A station-city integration evaluation index system covering multiple dimensions such as spatial structure, transportation connection, functional layout and operational efficiency is constructed. The indicators are normalized and the basic weight of each indicator is calculated by an objective weighting method based on the amount of data information. The spatial topological features, spatiotemporal dynamic features of passenger flow, and long-range dependency features between stations of the rail transit stations and their surrounding areas are extracted by using graph attention network, spatiotemporal graph convolutional network, and self-attention encoder, respectively, to obtain deep implicit features; The weighted sum of the basic weights and the normalized values ​​of the corresponding indicators is used to obtain the indicator weighted score. The deep implicit features are converted into deep feature scores through fully connected mapping. The indicator weighted score and the deep feature score are then weighted and fused to obtain the comprehensive score of station-city integration maturity. Based on the comprehensive score of station-city integration maturity, the station-city integration maturity level is output according to the preset level classification rules, and the comprehensive maturity score of multiple time phases is dynamically updated using time-series smoothing processing.

[0009] According to the method for evaluating the maturity of station-city integration provided by the present invention, the standardization preprocessing includes: The method involves mapping spatial data from different sources to a unified coordinate system, synchronizing data with different time frequencies using time references, and rasterizing discrete point data using inverse distance weighted interpolation.

[0010] According to the present invention, a method for evaluating the maturity of station-city integration is provided, wherein the evaluation index system includes: The indicators include land use mixing degree, which represents the degree of land use function mixing; functional density, which represents the level of facility agglomeration; road accessibility, which represents the degree of transportation convenience; and service coverage rate, which represents the coverage of public services.

[0011] According to the present invention, a method for evaluating the maturity of station-city integration is provided, wherein the objective weighting method based on the amount of data information is the entropy weight method. The basic weights of each indicator are calculated from the information entropy and redundancy of each indicator. The greater the degree of variation of an indicator, the higher its weight.

[0012] According to the method for evaluating the maturity of station-city integration provided by the present invention, the graph attention network is used to construct a heterogeneous graph of rail transit stations, land parcels and functional areas, and learn the spatial correlation importance between nodes through an attention mechanism; The spatiotemporal graph convolutional network is used to perform graph convolution operations on the time series of the OD passenger flow matrix to extract the spatiotemporal variation features of passenger flow. The self-attention encoder is used to capture global dependencies between different sites through a multi-head self-attention mechanism.

[0013] According to the method for evaluating the maturity of station-city integration provided by the present invention, the comprehensive score of station-city integration maturity is a linear weighted sum of the weighted scores of the indicators and the scores of the deep features, wherein the weighting coefficient is an adjustable parameter determined by the validation set.

[0014] According to the present invention, a method for evaluating the maturity of station-city integration is provided. The time-series smoothing process adopts an exponential weighted moving average method, which weights and sums the comprehensive maturity score of the current time phase with the smoothing score of the historical time phase according to a preset decay coefficient to obtain a dynamically updated comprehensive maturity score.

[0015] Secondly, the present invention also provides a station-city integration maturity evaluation system, comprising: The acquisition module is used to acquire multi-source heterogeneous data of rail transit stations and their surrounding areas and perform standardized preprocessing. The multi-source heterogeneous data includes rail transit passenger flow data, land use data, points of interest data, road network data, and population density data. The module is used to build a station-city integration evaluation index system covering multiple dimensions such as spatial structure, transportation connection, functional layout and operational efficiency. It normalizes each index and uses an objective weighting method based on data information to calculate the basic weight of each index. The extraction module is used to extract the spatial topological features, spatiotemporal dynamic features of passenger flow, and long-range dependency features between stations of the rail transit station and its surrounding area using graph attention network, spatiotemporal graph convolutional network, and self-attention encoder, respectively, to obtain deep hidden features. The weighting module is used to sum the basic weights and the normalized values ​​of the corresponding indicators to obtain the indicator weighted score, convert the deep implicit features into deep feature scores through fully connected mapping, and perform weighted fusion of the indicator weighted score and the deep feature score to obtain the comprehensive score of station-city integration maturity. The output module is used to output the maturity level of station-city integration based on the comprehensive score of station-city integration maturity according to the preset level classification rules, and to dynamically update the comprehensive maturity score of multiple time phases using time-series smoothing processing.

[0016] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement any of the above-described methods for evaluating the maturity of station-city integration.

[0017] Fourthly, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the station-city integration maturity evaluation method as described above.

[0018] The method and system for evaluating the maturity of station-city integration provided by this invention have the following beneficial effects: 1. The evaluation system has a high degree of standardization and covers a wide range of dimensions. This invention constructs a unified mathematical index system covering six dimensions: spatial structure, transportation connections, functional layout, operational efficiency, spatiotemporal behavior, and station-city coupling performance. This system overcomes the shortcomings of existing evaluation methods, such as fragmented dimensions, inconsistent calculation methods, and the inability to achieve cross-site horizontal comparisons. It is equipped with a standardized preprocessing workflow for multi-source heterogeneous data, which can complete the systematic integration and processing of various types of data, such as passenger flow, land use, POI, road network, and population density. This supports the full-dimensional quantitative evaluation of station-city integration, and the evaluation results are more scientific and reliable.

[0019] 2. The indicator weights are objective and stable, avoiding subjective experience bias. This invention adopts a weight generation strategy of "entropy weight method basic calculation + machine learning feature importance calibration". It completes the objective calculation and optimization of weights entirely based on the characteristics of the data itself, replacing the traditional weight determination methods that rely on subjective experience, such as manual setting and analytic hierarchy process. This effectively reduces the evaluation bias caused by human factors and improves the stability and reproducibility of the evaluation results.

[0020] 3. Strong ability to characterize complex coupling features and high level of intelligence. This invention introduces three types of deep learning models: Graph Attention Network (GAT), Spatiotemporal Graph Convolutional Network (ST-GCN), and TransformerEncoder. These models can automatically extract nonlinear implicit features such as spatial topological relationships of station areas, spatiotemporal dynamic changes in passenger flow, and long-range dependencies between stations. This makes up for the shortcomings of traditional static evaluation methods in characterizing the complex coupling relationship between stations and cities, and significantly improves the adaptability of the evaluation method to complex urban systems and the accuracy of the evaluation results.

[0021] 4. Possesses dynamic updating capabilities to match urban development patterns. This invention employs the Exponential Weighted Moving Average (EWMA) method to construct a dynamic and smooth update mechanism for maturity, which can achieve iterative updates of the station-city integration level based on time-series data. This overcomes the shortcomings of existing static evaluation methods that cannot capture the dynamic evolution of the functions and vitality around the station. The evaluation results can truly reflect the real-time development status of the station and meet the needs of dynamic urban governance.

[0022] 5. The solution is flexible and scalable, and highly adaptable to various scenarios. The evaluation scheme of this invention has multi-dimensional adjustability: the maturity level classification threshold can be flexibly adjusted according to the city's development stage and application scenario requirements; the evaluation indicators can be expanded as needed according to the evaluation object; and the deep learning model can be replaced according to data characteristics. It can adapt to the evaluation needs of different cities, different routes, and different types of stations, and has a wide range of applications.

[0023] 6. Fully automated closed-loop process, highly practical for implementation. This invention is accompanied by a complete modular evaluation system, forming a fully automated processing chain from data access, indicator calculation, deep learning modeling, maturity determination to visualization output. It solves the problems of excessive manual intervention, low execution efficiency, and poor reusability in existing evaluation processes. The accompanying visualization module supports functions such as map spatial display, maturity time-series tracking, and horizontal comparison of multiple stations. It can directly provide quantitative decision support for rail transit planning, TOD assessment, urban renewal, and refined station operation, and has high practical value and promotion significance. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0025] Figure 1 This is one of the flowcharts of the station-city integration maturity evaluation method provided by the present invention; Figure 2This is the second flowchart of the station-city integration maturity evaluation method provided by the present invention; Figure 3 This is a schematic diagram of the structure of the station-city integration maturity evaluation system provided by the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0027] This invention addresses the technical problems in existing station-city integration evaluation systems, such as the lack of unified standards in the indicator system, strong subjectivity in weight setting, insufficient multi-source data fusion capabilities, weak ability to characterize complex spatiotemporal features, static evaluation models, and low automation. It provides a method for evaluating the maturity of station-city integration based on multi-source data and deep learning, along with a supporting evaluation system and computer-readable storage media. The solution uses multi-source heterogeneous urban data as a foundation to construct a standardized multi-dimensional evaluation indicator system. It combines entropy weighting and feature importance to achieve objective weight calibration, introduces three types of deep learning models to extract spatial, temporal, and semantic implicit features, and calculates the maturity score through a dual-channel fusion of "indicator weighting + deep features." With a supporting dynamic update mechanism and a fully automated system, it achieves an objective, quantifiable, scalable, and dynamically updatable intelligent evaluation of station-city integration maturity.

[0028] Figure 1 This is one of the flowcharts illustrating the station-city integration maturity evaluation method provided in this embodiment of the invention, such as... Figure 1 As shown, it includes: Step 100: Acquire multi-source heterogeneous data of rail transit stations and their surrounding areas and perform standardized preprocessing. The multi-source heterogeneous data includes rail transit passenger flow data, land use data, points of interest data, road network data, and population density data. Step 200: Construct a station-city integration evaluation index system covering multiple dimensions such as spatial structure, transportation connection, functional layout and operational efficiency, normalize each index, and use an objective weighting method based on data information to calculate the basic weight of each index. Step 300: Extract the spatial topological features, spatiotemporal dynamic features of passenger flow, and long-range dependency features between stations of the rail transit station and its surrounding area using graph attention network, spatiotemporal graph convolutional network, and self-attention encoder respectively, to obtain deep implicit features; Step 400: The basic weights and the normalized values ​​of the corresponding indicators are weighted and summed to obtain the indicator weighted score. The deep latent features are converted into deep feature scores through fully connected mapping. The indicator weighted score and the deep feature score are weighted and fused to obtain the comprehensive score of station-city integration maturity. Step 500: Based on the comprehensive score of station-city integration maturity, output the station-city integration maturity level according to the preset level classification rules, and use time-series smoothing processing to dynamically update the comprehensive maturity score of multiple time phases.

[0029] Specifically, such as Figure 2 As shown, the embodiments of the present invention include the following steps: Step 1: Multi-source data acquisition and preprocessing Collect heterogeneous data from multiple sources, including rail transit passenger flow data, land use data, POI data, road network data, and population density data, and perform three types of standardized preprocessing: (1) Coordinate transformation Mapping data from different sources to a unified coordinate system:

[0030] (2) Time synchronization processing Establish a unified time base for data from different time frequencies:

[0031] (3) Spatial interpolation (IDW) Rasterization for population density, POI intensity, etc.

[0032] Step 2: Constructing a multi-dimensional indicator system for station-city integration Establish a standardized evaluation index system covering different dimensions, with corresponding quantifiable and calculable indicators set for each dimension: (1) Land use mixing index

[0033] (2) Functional Density (POI Density)

[0034] (3) Road accessibility index

[0035] (4) Service coverage

[0036] Step 3: Indicator Normalization and Weight Learning (1) Min-Max Normalization

[0037] (2) Entropy weight method Indicator percentage:

[0038] Information entropy:

[0039] Redundancy:

[0040] Weight:

[0041] Step 4: Deep Learning Feature Extraction Model Based on the indicator system, this invention further introduces a deep learning model to automatically extract the implicit structural features of station-city integration, thereby improving the ability to characterize complex spatial behaviors and spatiotemporal patterns.

[0042] This embodiment uses three types of models: (1) Graph Attention Network (GAT); (2) Spatiotemporal Graph Convolutional Network (ST-GCN); and (3) Transformer Encoder.

[0043] (1) Graph Attention Network (GAT) for Spatial Topology Modeling The sites, plots, and functional zones are constructed as heterogeneous graphs. The importance of nodes is learned through attention mechanisms.

[0044]

[0045] The attention coefficient is calculated as follows:

[0046] (2) Spatiotemporal graph convolutional network (ST-GCN) modeling the spatiotemporal characteristics of OD flow Convolution is performed on the time series passenger flow matrix Ht:

[0047] (3) Transformer Encoder extracts high-dimensional semantic features Use a self-attention mechanism to capture long-range dependencies between sites:

[0048] in:

[0049] Step 5: Comprehensive Calculation of Integration Maturity This invention adopts a dual-channel fusion approach of "indicator system + deep features".

[0050] (1) Linear fusion scoring of indicator system

[0051] (2) Deep feature score transformation Deep features are processed through a fully connected layer. Mapped to rating:

[0052] (3) Overall integration maturity score

[0053] in It is an adjustable parameter (0.3~0.7), determined by the validation set according to the task.

[0054] Step 6: Maturity Level Classification For ease of application, the maturity score is discretized into six levels:

[0055] Each level represents a different degree of integration between the station and the city: Level 0: Extremely low integration. Spatially fragmented, with weak transportation connections, limited functionality, and a lack of urban activity support, making it difficult to attract visitors; Level 1: Low integration. Basic connectivity exists, but the connection efficiency is low; the functional layout does not match the traffic flow; and the intensity of urban activities is low. Level 2: Low to medium integration. It possesses some traffic attraction capacity, but its functional structure is relatively weak, and the surrounding area lacks active updates. Level 3: Moderate integration. Functional areas are gradually being improved, transportation convenience is significantly enhanced, and urban vitality is beginning to gather; Level 4: High integration. Transportation, function, and space are efficiently coordinated, exhibiting a high degree of vitality and functional integration, and possessing typical TOD demonstration characteristics; Level 5: Extremely high integration. The station and city are deeply integrated, with excellent performance across multiple dimensions. It has formed a mature and highly vibrant urban node and can serve as a best practice example.

[0056] To avoid arbitrary threshold determinations, this invention incorporates the following technical basis: (1) Statistical Binning K-means clustering or equal-frequency binning is performed on the S values ​​of a large number of sites to obtain natural segments.

[0057] (2) Expert experience and industry standards Refer to TOD evaluation systems (such as cases from mature cities like Japan) to determine reference thresholds.

[0058] (3) Model error minimization principle Cross-validation minimizes the classification error after dividing the rank intervals.

[0059] The grading threshold of this invention is not fixed and can be adjusted according to factors such as the stage of urban development, changes in data statistical distribution, characteristics of different lines / stations, and application needs of users (government / planning departments).

[0060] Therefore, the grading function can be expanded to:

[0061] Step 7: Spatiotemporal Dynamic Update Considering the smoothness of time series data, this invention employs an exponentially weighted moving average (EWMA):

[0062] in, .

[0063] Step 8: Implementation Examples Taking 58 rail transit stations in a certain city as an example, the maturity score is calculated based on the method of this invention:

[0064] Based on the grading function:

[0065] The result is "high integration," which can support planning decisions and operational optimization.

[0066] The following describes the station-city integration maturity evaluation system provided by the present invention. The station-city integration maturity evaluation system described below and the station-city integration maturity evaluation method described above can be referred to in correspondence.

[0067] Figure 3 This is a schematic diagram of the structure of the station-city integration maturity evaluation system provided in an embodiment of the present invention, as shown below. Figure 3 As shown, it includes: an acquisition module 31, a construction module 32, an extraction module 33, a weighting module 34, and an output module 35, wherein: The acquisition module 31 is used to acquire multi-source heterogeneous data of rail transit stations and their surrounding areas and perform standardized preprocessing. The multi-source heterogeneous data includes rail transit passenger flow data, land use data, points of interest data, road network data, and population density data. The construction module 32 is used to construct a station-city integration evaluation index system covering multiple dimensions such as spatial structure, transportation connection, functional layout, and operational efficiency. It normalizes each index and calculates the basic weight of each index using an objective weighting method based on data information volume. The extraction module 33 is used to extract data from the rail transit stations and their surrounding areas using graph attention networks, spatiotemporal graph convolutional networks, and self-attention encoders, respectively. The spatial topological features of the domain, the spatiotemporal dynamic features of passenger flow, and the long-range dependency features between stations are extracted to obtain deep implicit features. The weighting module 34 is used to weight and sum the basic weights with the normalized values ​​of the corresponding indicators to obtain the indicator weighted score. The deep implicit features are converted into deep feature scores through fully connected mapping. The indicator weighted scores and the deep feature scores are weighted and fused to obtain the comprehensive score of station-city integration maturity. The output module 35 is used to output the station-city integration maturity level according to the preset level division rules based on the comprehensive score of station-city integration maturity, and to dynamically update the comprehensive maturity scores of multiple time phases using time-series smoothing processing.

[0068] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440. The processor 410, communication interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a station-city integration maturity evaluation method. This method includes: acquiring multi-source heterogeneous data of rail transit stations and their surrounding areas and performing standardized preprocessing; the multi-source heterogeneous data includes rail transit passenger flow data, land use data, points of interest data, road network data, and population density data; constructing a station-city integration evaluation index system covering multiple dimensions such as spatial structure, transportation connections, functional layout, and operational efficiency; normalizing each index; and calculating the basic weight of each index using an objective weighting method based on data information volume; and utilizing graph attention networks, spatiotemporal graph convolutional networks, and self-attention networks respectively. The force encoder extracts the spatial topological features, spatiotemporal dynamic features of passenger flow, and long-range dependency features between stations of the rail transit station and its surrounding area to obtain deep implicit features. The basic weights are weighted and summed with the normalized values ​​of the corresponding indicators to obtain the indicator weighted score. The deep implicit features are converted into deep feature scores through fully connected mapping. The indicator weighted scores and the deep feature scores are weighted and fused to obtain the comprehensive score of station-city integration maturity. Based on the comprehensive score of station-city integration maturity, the station-city integration maturity level is output according to the preset level classification rules, and the maturity comprehensive score of multiple time phases is dynamically updated using temporal smoothing processing.

[0069] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0070] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the station-city integration maturity evaluation method provided by the above-described methods. This method includes: acquiring multi-source heterogeneous data of rail transit stations and their surrounding areas and performing standardized preprocessing; the multi-source heterogeneous data includes rail transit passenger flow data, land use data, points of interest data, road network data, and population density data; constructing a station-city integration evaluation index system covering multiple dimensions such as spatial structure, transportation connections, functional layout, and operational efficiency; normalizing each index; and calculating the basic weight of each index using an objective weighting method based on data information volume; and respectively utilizing... The spatial topological features, spatiotemporal dynamic features of passenger flow, and long-range dependency features between stations of the rail transit stations and their surrounding areas are extracted using graph attention networks, spatiotemporal graph convolutional networks, and self-attention encoders to obtain deep latent features. The basic weights are weighted and summed with the normalized values ​​of the corresponding indicators to obtain the indicator weighted scores. The deep latent features are converted into deep feature scores through fully connected mapping. The indicator weighted scores and the deep feature scores are then weighted and fused to obtain the comprehensive station-city integration maturity score. Based on the comprehensive station-city integration maturity score, the station-city integration maturity level is output according to a preset level classification rule, and the comprehensive maturity score for multiple time phases is dynamically updated using temporal smoothing processing.

[0071] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0072] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating the maturity of station-city integration, characterized in that, include: Acquire multi-source heterogeneous data of rail transit stations and their surrounding areas and perform standardized preprocessing. The multi-source heterogeneous data includes rail transit passenger flow data, land use data, points of interest data, road network data, and population density data. A station-city integration evaluation index system covering multiple dimensions such as spatial structure, transportation connection, functional layout and operational efficiency is constructed. The indicators are normalized and the basic weight of each indicator is calculated by an objective weighting method based on the amount of data information. The spatial topological features, spatiotemporal dynamic features of passenger flow, and long-range dependency features between stations of the rail transit stations and their surrounding areas are extracted by using graph attention network, spatiotemporal graph convolutional network, and self-attention encoder, respectively, to obtain deep implicit features; The weighted sum of the basic weights and the normalized values ​​of the corresponding indicators is used to obtain the indicator weighted score. The deep implicit features are converted into deep feature scores through fully connected mapping. The indicator weighted score and the deep feature score are then weighted and fused to obtain the comprehensive score of station-city integration maturity. Based on the comprehensive score of station-city integration maturity, the station-city integration maturity level is output according to the preset level classification rules, and the comprehensive maturity score of multiple time phases is dynamically updated using time-series smoothing processing.

2. The method for evaluating the maturity of station-city integration according to claim 1, characterized in that, The standardized preprocessing includes: The method involves mapping spatial data from different sources to a unified coordinate system, synchronizing data with different time frequencies using time references, and rasterizing discrete point data using inverse distance weighted interpolation.

3. The method for evaluating the maturity of station-city integration according to claim 1, characterized in that, The evaluation index system includes: The indicators include land use mixing degree, which represents the degree of land use function mixing; functional density, which represents the level of facility agglomeration; road accessibility, which represents the degree of transportation convenience; and service coverage rate, which represents the coverage of public services.

4. The method for evaluating the maturity of station-city integration according to claim 1, characterized in that, The objective weighting method based on data information content is the entropy weighting method; The basic weights of each indicator are calculated from the information entropy and redundancy of each indicator. The greater the degree of variation of an indicator, the higher its weight.

5. The method for evaluating the maturity of station-city integration according to claim 1, characterized in that, The graph attention network is used to construct a heterogeneous graph of rail transit stations, plots and functional areas, and learns the spatial relationship importance between nodes through an attention mechanism. The spatiotemporal graph convolutional network is used to perform graph convolution operations on the time series of the OD passenger flow matrix to extract the spatiotemporal variation features of passenger flow. The self-attention encoder is used to capture global dependencies between different sites through a multi-head self-attention mechanism.

6. The method for evaluating the maturity of station-city integration according to claim 1, characterized in that, The comprehensive score for the maturity of station-city integration is a linear weighted sum of the weighted scores of the indicators and the scores of the deep features, wherein the weighting coefficient is an adjustable parameter determined by the validation set.

7. The method for evaluating the maturity of station-city integration according to claim 1, characterized in that, The temporal smoothing process employs an exponentially weighted moving average method, which weights and sums the current maturity score with the smoothing scores of historical time phases using a preset decay coefficient to obtain a dynamically updated maturity score.

8. A system for evaluating the maturity of station-city integration, characterized in that, include: The acquisition module is used to acquire multi-source heterogeneous data of rail transit stations and their surrounding areas and perform standardized preprocessing. The multi-source heterogeneous data includes rail transit passenger flow data, land use data, points of interest data, road network data, and population density data. The module is used to build a station-city integration evaluation index system covering multiple dimensions such as spatial structure, transportation connection, functional layout and operational efficiency. It normalizes each index and uses an objective weighting method based on data information to calculate the basic weight of each index. The extraction module is used to extract the spatial topological features, spatiotemporal dynamic features of passenger flow, and long-range dependency features between stations of the rail transit station and its surrounding area using graph attention network, spatiotemporal graph convolutional network, and self-attention encoder, respectively, to obtain deep hidden features. The weighting module is used to sum the basic weights and the normalized values ​​of the corresponding indicators to obtain the indicator weighted score, convert the deep implicit features into deep feature scores through fully connected mapping, and perform weighted fusion of the indicator weighted score and the deep feature score to obtain the comprehensive score of station-city integration maturity. The output module is used to output the maturity level of station-city integration based on the comprehensive score of station-city integration maturity according to the preset level classification rules, and to dynamically update the comprehensive maturity score of multiple time phases using time-series smoothing processing.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the station-city integration maturity evaluation method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the station-city integration maturity evaluation method as described in any one of claims 1 to 7.

Images