A customized public transport intelligent line arrangement method, system and device based on line opinion collection and a storage medium

Abstract

This application provides a method, system, device, and storage medium for intelligent route planning of customized buses based on route opinion collection. During the clustering process, this method adaptively matches and sets core clustering parameters according to the regional functions, travel times, and data scale of the demand points. Simultaneously, it calculates the comprehensive similarity between demand points by combining spatial and temporal correlation weights, constructing a corresponding similarity matrix to complete clustering. This achieves spatiotemporal dual-dimensional fusion clustering, adapting to the travel demand characteristics of different scenarios and time periods, ensuring that the clustering results closely match users' actual travel patterns. Secondly, this method adopts a multi-objective collaborative optimization architecture, simultaneously considering path cost, vehicle fixed cost, and passenger satisfaction, avoiding the deterioration of operational indicators caused by single-objective optimization. The generated route plans are more in line with the needs of commercial operation of customized buses.

Description

Technical Field

[0001] This application relates to the field of intelligent public transportation planning technology, specifically to a customized intelligent bus route planning method, system, equipment, and storage medium based on route opinion collection. Background Technology

[0002] With the continuous growth of personalized travel demands from urban residents, customized bus services have become an important supplement to the public transportation system. The core of customized bus services is to aggregate stops and plan routes based on users' submitted pick-up and drop-off locations and travel time requirements, generating bus route solutions that balance operating costs, service quality, and traffic feasibility.

[0003] Current customized bus technologies are mostly based on the traditional DBSCAN spatial clustering algorithm, whose core process is: simple data cleaning, fixed radius spatial clustering, and greedy path connection.

[0004] However, existing technologies have the following shortcomings in the preprocessing and clustering of demand data: 1) Clustering based solely on static spatial location fails to capture seasonal demand fluctuations such as morning and evening rush hours, school opening seasons, and holidays, resulting in a severe disconnect between clustering results and users' actual travel patterns; 2) It cannot adapt to the differences between urban and rural scenarios. Urban demand is more concentrated, which can easily lead to under-clustering, while rural demand is more sparse, which can easily lead to over-segmentation; 3) Data cleaning only involves simple outlier removal without integrating data such as road network and weather for correction. Invalid or erroneous data can directly interfere with clustering accuracy, causing subsequent route planning to deviate from actual travel needs.

[0005] Secondly, existing technologies have the following drawbacks in generating customized bus routes: 1) Existing technologies mostly adopt a single-objective optimization architecture, focusing solely on minimizing the total route distance. This easily leads to problems such as shortened routes but a doubling of transfers and soaring empty-load rates, resulting in a severe deterioration of actual operational indicators and an inability to meet the commercial operation requirements of customized buses. 2) Existing route modeling often uses traditional simple graph structures, only achieving pairwise relationships between nodes. The generated routes only achieve theoretical path optimization, completely detached from actual road conditions and unforeseen environmental constraints, making them unsuitable for practical operation. 3) The weight coefficients of existing optimization models rely entirely on manual experience for configuration, resulting in extremely poor adaptability to large-scale complex road networks. This leads to slow convergence speed, solution quality heavily influenced by initial parameters, and significant fluctuations in optimization performance across different scenarios.

[0006] Therefore, how to design a customized intelligent bus route planning method, system, equipment, and storage medium based on route opinion collection that can adapt to time-varying demand fluctuations and urban-rural differences, while taking into account route distance and commercial needs, is a technical problem that has not yet been solved in the existing technology. Summary of the Invention

[0007] Therefore, this application proposes a customized intelligent bus route planning method, system, equipment, and storage medium based on route opinion collection that can adapt to time-varying demand fluctuations and urban-rural scenario differences, while taking into account route distance and commercial needs, in order to improve or solve at least one of the above-mentioned technical problems.

[0008] This application mainly includes the following aspects: In a first aspect, embodiments of this application provide a customized intelligent bus route planning method based on route feedback collection, the method comprising: S101: Obtain customized public transport travel demand data and matching multimodal environmental data to construct an original spatiotemporal demand database; S102: Based on the multimodal environment data, perform outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database to obtain a cleaned spatiotemporal demand dataset. S103: Standardize the cleaned spatiotemporal demand dataset, and construct a standardized spatiotemporal object for the boarding point and a standardized spatiotemporal object for the alighting point with a common unique demand identifier for each valid demand data, forming a set of standardized spatiotemporal objects for the boarding point and a set of standardized spatiotemporal objects for the alighting point respectively. S104: Based on the regional function, travel time period and data scale of the standardized spatiotemporal object of the boarding point, adaptively match and set the core parameters of clustering; Based on the standardized spatiotemporal object set of the boarding point and the clustering core parameters, calculate the comprehensive similarity between any two standardized spatiotemporal objects of the boarding point and construct a comprehensive similarity matrix of the boarding point. S105: Based on the calibrated boarding point comprehensive similarity matrix and the minimum number of cluster points in the clustering core parameters, traverse, judge and filter the spatiotemporal core objects and potential noise points of the boarding points; iteratively generate the spatiotemporal demand clusters of the boarding points based on the spatiotemporal core objects of the boarding points, and after completing the secondary verification and filtering of the potential noise points, obtain the set of effective spatiotemporal demand clusters of the boarding points. S106: Based on the standardized spatiotemporal object set of the drop-off point and the clustering core parameters, replace the execution subject in S104 to S105 with the standardized spatiotemporal object of the drop-off point and execute again to obtain a set of effective spatiotemporal demand clusters of drop-off points. S107: Complete full-dimensional attribute encapsulation for all the valid boarding point spatiotemporal demand clusters and the valid alighting point spatiotemporal demand clusters respectively, to obtain fully encapsulated boarding point spatiotemporal demand clusters and fully encapsulated alighting point spatiotemporal demand clusters, thereby forming a fully encapsulated boarding point spatiotemporal demand cluster set and a fully encapsulated alighting point spatiotemporal demand cluster set. Both the fully enclosed boarding point spatiotemporal demand cluster set and the fully enclosed disembarkation point spatiotemporal demand cluster set include: spatial attributes, temporal attributes, demand intensity attributes, and association attributes; wherein, the spatial attributes include: cluster center coordinates, geographical coverage boundary, and a list of associated road nodes; the temporal attributes include: a set of timestamps and time distribution characteristics; the demand intensity attributes include: cluster demand quantity, demand density, and priority weight; and the association attributes include original demand data reference links and environmental event tags; S108: Based on the unique demand identifier, the set of spatiotemporal demand clusters of fully enclosed boarding points and the set of spatiotemporal demand clusters of fully enclosed alighting points, construct the boarding point cluster-alighting point cluster OD demand matrix, and output the set of spatiotemporal demand clusters of fully enclosed boarding points, the set of spatiotemporal demand clusters of fully enclosed alighting points and the set of boarding point cluster-alighting point cluster OD demand matrix that can be used for route planning. S109: Based on the fully enclosed boarding point spatiotemporal demand cluster set, the fully enclosed alighting point spatiotemporal demand cluster set, the boarding point cluster-alighting point cluster OD demand matrix, and the supporting basic data for route planning, construct a basic dataset for customized bus route planning; wherein, the fully enclosed boarding point spatiotemporal demand cluster set, the fully enclosed alighting point spatiotemporal demand cluster set, and the boarding point cluster-alighting point cluster OD demand matrix are merged into a spatiotemporal demand cluster set; the supporting basic data for route planning includes urban road network topology data, road construction and road closure event data, severe weather data, road segment toll cost data, historical traffic flow data, and bus vehicle operation parameters associated with the spatiotemporal demand clusters. S110: Based on the customized bus route planning basic dataset, define demand nodes, road nodes, environmental event nodes and hyperedge mapping rules, construct an initial hypergraph topology, and realize cross-dimensional unified association of demand, road network and environment; S111: By constructing a dynamic weight function for each hyperedge and assigning weights, a weighted dynamic hypergraph structure is generated, enabling a quantitative representation of the priority and passage constraints of different demand clusters. S112: Based on the weighted dynamic hypergraph structure, a two-layer graph convolutional network is constructed and inter-layer convolution is performed by inputting the hypergraph adjacency matrix and the node attribute matrix to extract the spatial association, demand intensity and environmental constraint features of the nodes, and generate a set of hypergraph node embedding vectors. S113: Based on the set of embedded vectors of the hypergraph nodes, a meta-learning network containing two fully connected layers and a normalized exponential function layer is constructed to complete the nonlinear mapping and normalization of node features to multi-objective weights, and output dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights, thereby realizing the adaptive dynamic allocation of multi-objective optimization weights. S114: Based on the customized bus route planning basic dataset, the weighted dynamic hypergraph structure, and the dynamic weight coefficients, construct a multi-objective optimization overall objective function and inject hard constraints to form a complete multi-objective mixed integer programming model; S115: Use an integer programming solver to solve the multi-objective mixed integer programming model to obtain multiple initial route schemes. Calculate the core operation indicators of each initial route scheme and complete compliance verification. After eliminating invalid schemes, output a pool of customized bus candidate route schemes.

[0009] According to one embodiment of the present invention, the steps of acquiring customized public transport travel demand data and matching multimodal environmental data to construct an original spatiotemporal demand database, and performing outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database based on the multimodal environmental data to obtain a cleaned spatiotemporal demand dataset include: The customized bus travel demand data is obtained through the user terminal. Each piece of customized bus travel demand data includes the unique demand identifier, the latitude and longitude coordinates of the boarding point, the latitude and longitude coordinates of the alighting point, the user's expected boarding time window, and the user's expected alighting time window. Acquire the multimodal environment data that matches the customized public transportation travel demand data. The multimodal environment data includes urban road network topology data, weather warning data, historical traffic status data, and special event identifier data for holidays and school opening seasons. All the acquired customized public transport travel demand data are correlated and matched with the multimodal environment data in terms of time and space. The user's expected boarding time window and the user's expected alighting time window are used as the time reference, and the boarding point latitude and longitude coordinates and the alighting point latitude and longitude coordinates are used as the spatial reference. The corresponding spatiotemporal range of the multimodal environment data is bound to each customized public transport travel demand data to construct the original spatiotemporal demand database. Perform anomaly removal operation: Based on the urban road network topology data, calculate the vertical distance from the latitude and longitude coordinates of each boarding point and the latitude and longitude coordinates of each alighting point to the nearest road network, and remove data with a vertical distance exceeding 500 meters. Perform duplicate request data deduplication: For multiple identical request data submitted by the same user, only one valid data is retained, and all other duplicate data is removed; Perform demand time window adjustment operation: Based on the weather warning data, if severe weather exists in the corresponding time and space range, the start time of the user's expected boarding time window will be extended forward by 10% and the end time of the user's expected alighting time window will be extended backward by 10%. Based on the historical traffic data, if the corresponding spatiotemporal range belongs to a normally congested road section, the start time of the user's expected boarding time window will be extended forward by 15% and the end time of the user's expected alighting time window will be extended backward by 15%. Based on the special event identifier data of holidays and school opening season, if the corresponding spatiotemporal range is in the peak travel period of holidays or the school opening season, the start time of the user's expected boarding time window will be extended forward by 20% and the end time of the user's expected alighting time window will be extended backward by 20%. If a single customized bus travel demand data point simultaneously meets multiple of the above correction conditions, the maximum extension duration among the conditions will be taken, without repeated addition of durations; the upper limit threshold for the extension duration is 30 minutes.

[0010] According to one embodiment of the present invention, the structure of the standardized spatiotemporal object of the boarding point is as follows: ; in, This is the unique requirement identifier. Here are the longitude coordinates of the boarding point. Here are the latitude coordinates of the boarding point. This is the midpoint timestamp of the user's expected boarding time window after the demand time window has been corrected. The structure of the standardized spatiotemporal object of the drop-off point is as follows: ; in, The unique demand identifier is shared with the standardized spatiotemporal object of the boarding point. Here are the longitude coordinates of the drop-off point. Here are the latitude coordinates of the drop-off point. The timestamp of the midpoint of the user's expected drop-off time window after correction by the demand time window; The core clustering parameters include the spatial decay radius. Time decay coefficient and minimum number of cluster points The steps for adaptive matching settings of the clustering core parameters include: Based on the functional areas of the standardized spatiotemporal objects at the boarding points, they are classified into commercial and industrial parks, residential areas, and rural areas, and corresponding spatial attenuation radii are set for each. ; Based on the standardized spatiotemporal objects of the boarding points, the travel time periods are classified into weekday morning and evening peak hours, weekday off-peak hours, and holidays, and corresponding time decay coefficients are set for each. ; Based on the data scale of the standardized spatiotemporal object of the boarding point, the minimum number of cluster points is calculated. : ; in, The total number of standardized spatiotemporal objects for the boarding point within the standardized spatiotemporal object set for the boarding point. This represents the total number of pre-set customized bus routes. Based on the standardized spatiotemporal object set of boarding points and the clustering core parameters, the steps for calculating the comprehensive similarity between any two standardized spatiotemporal objects of boarding points and constructing the comprehensive similarity matrix of boarding points include: Iterate through all pairwise combinations of the standardized spatiotemporal objects of the boarding points within the set of standardized spatiotemporal objects of the boarding points. and Calculate the spatial correlation weight between the two points respectively. Time-related weights ; Wherein, the spatial association weight : ; in, Standardize the spatiotemporal object of the boarding point and The Euclidean distance is calculated using latitude and longitude coordinates; The spatial attenuation radius; The time-related weight : ; in, The absolute difference between the timestamps of the two points; The time decay coefficient is mentioned above; Based on the spatial correlation weight Time-related weights Calculate the comprehensive similarity : ; The overall similarity The range of values ​​is A higher value indicates a stronger spatial proximity and temporal consistency between the two points, and a higher cluster correlation. Complete the comprehensive similarity of all pairwise combinations of spatiotemporal objects within the standardized spatiotemporal object set of the boarding point. After calculation, a comprehensive similarity matrix for the boarding points is constructed. This matrix is ​​a two-dimensional square matrix with equal rows and columns. Each row and column corresponds to a standardized spatiotemporal object of the boarding point, and each cell element represents the comprehensive similarity score. .

[0011] According to one embodiment of the present invention, the steps of traversing, judging, and filtering the spatiotemporal core objects and potential noise points of the boarding points based on the calibrated comprehensive similarity matrix of the boarding points and the minimum number of cluster points in the clustering core parameters; iteratively generating spatiotemporal demand clusters of boarding points based on the spatiotemporal core objects of the boarding points; and obtaining a set of effective spatiotemporal demand clusters of boarding points after completing the secondary verification and filtering of the potential noise points include: S1051: Initialize the classification state of all the standardized spatiotemporal objects at the boarding points to unclassified and the processing state to unprocessed; initialize the initial value of the auto-incrementing cluster number sequence to 0; and initialize the set of potential noise points to empty. The classification status includes: unclassified, classified, and marked; where unclassified means that the object has not completed the verification and attribution determination; classified means that the object has been assigned to the determined spatiotemporal demand cluster of the boarding point; marked means that the object has completed the first round of verification and has been identified as a potential noise point, and is waiting for the second verification and filtering. S1052: Traverse each of the standardized spatiotemporal objects of the boarding point in the set of standardized spatiotemporal objects of the boarding point and check its classification status: if it is classified, continue to traverse the next standardized spatiotemporal object of the boarding point; if it is unclassified, treat it as an unclassified spatiotemporal object Pi and execute S1053. S1053: For the unclassified spatiotemporal object Pi, select the comprehensive similarity scores with the unclassified spatiotemporal object Pi from the comprehensive similarity matrix of the boarding point. All standardized spatiotemporal objects of the boarding points that are greater than or equal to the neighborhood similarity threshold constitute the spatiotemporal neighborhood point set of the unclassified spatiotemporal object Pi; the neighborhood similarity threshold is a preset configurable parameter; S1054: Count the total number of standardized spatiotemporal objects of the boarding point within the spatiotemporal neighborhood point set. If the total number is less than the minimum clustering point number... If the unclassified spatiotemporal object Pi is determined to be a potential noise point, it is added to the potential noise point set, and its classification status is updated to marked. Then, return to S1052 to continue the traversal operation. If the total number is greater than or equal to the minimum number of cluster points If so, the unclassified spatiotemporal object Pi is determined to be the core spatiotemporal object of the boarding point, the auto-incrementing cluster number is assigned to it, and the process proceeds to S1055. S1055: All the core spatiotemporal objects of the boarding point, as well as all the standardized spatiotemporal objects of the boarding point in its spatiotemporal neighborhood point set, are classified into the spatiotemporal demand cluster of the boarding point corresponding to the auto-incrementing cluster number, and their classification status is uniformly updated to classified. S1056: Iterate through all standardized spatiotemporal objects of the current boarding point spatiotemporal demand cluster whose processing status is unprocessed. For each currently traversed cluster object Px, filter out the comprehensive similarity between the cluster object Px and the current cluster object Px from the comprehensive similarity matrix of the boarding points. All standardized spatiotemporal objects of the boarding points that are greater than or equal to the neighborhood similarity threshold constitute the spatiotemporal neighborhood point set of the intra-cluster object Px; Count the total number of objects in the spatiotemporal neighborhood point set of the cluster object Px. If the total number is greater than or equal to the minimum clustering point number... If the object Px in the cluster is determined to be the new spatiotemporal core object of the boarding point, all objects in the spatiotemporal neighborhood point set whose classification status is unclassified are simultaneously assigned to the spatiotemporal demand cluster of the boarding point corresponding to the current auto-incrementing cluster number, and their classification status is updated to classified. After completing the above judgment and attribution operations, the processing status of the current cluster object Px is updated to processed; the above traversal operation is continued until the processing status of all objects in the boarding point spatiotemporal demand cluster is processed, thereby completing the generation of the boarding point spatiotemporal demand cluster, and the value of the auto-incrementing cluster number is incremented by one, and then proceeding to S1057. S1057: If there are still unclassified standardized spatiotemporal objects in the set of standardized spatiotemporal objects of the boarding point, return to S1052 to continue the traversal operation; if all the standardized spatiotemporal objects of the boarding point are classified or marked, proceed to S1058. S1058: For all potential noise points in the set of potential noise points, perform secondary verification and filtering: calculate the comprehensive similarity between the potential noise point and all generated spatiotemporal core objects of the boarding point. Take the maximum value among them; If the maximum value is greater than or equal to the noise point attribution threshold, then the potential noise point is assigned to the spatiotemporal demand cluster of the boarding point spatiotemporal core object corresponding to the maximum value, and its classification status is updated to classified; the noise point attribution threshold is a preset configurable parameter and is less than the neighborhood similarity threshold. If the maximum value is less than the noise point attribution threshold, the potential noise point is marked as a spatiotemporally isolated noise point and is not included in any of the boarding point spatiotemporal demand clusters. S1059: After completing the generation of all the aforementioned spatiotemporal demand clusters for boarding points and the filtering of noise points, the set of effective spatiotemporal demand clusters for boarding points is obtained.

[0012] According to one embodiment of the present invention, based on the customized bus route planning basic dataset, demand nodes, road nodes, environmental event nodes, and hyperedge mapping rules are defined to construct an initial hypergraph topology structure, realizing cross-dimensional unified association of demand, road network, and environment. The steps of generating a weighted dynamic hypergraph structure by constructing a dynamic weight function for each hyperedge and assigning weights to it, and realizing the quantitative representation of the priority and traffic constraints of different demand clusters, include: Define the demand node, the road node, and the environmental event node; wherein, the demand node corresponds to the spatiotemporal demand cluster, the road node corresponds to the intersection and road segment endpoint in the urban road network topology data, and the environmental event node corresponds to the road construction and road closure event data and the event node affecting traffic in the severe weather data; Each demand node is mapped to an independent hyperedge, and each hyperedge connects all the road nodes covered by the virtual site corresponding to the demand node, as well as all the environmental event nodes within the time distribution range of the demand node, thereby realizing a cross-dimensional unified association between demand, road network and environment. Based on the definitions of the demand nodes, road nodes, and environmental event nodes, as well as the hyperedge mapping rules, the initial hypergraph topology is constructed, clarifying the association relationships of all nodes and the coverage of hyperedges; Construct the dynamic weight function, which is as follows: ; in, For the current planning period; The base travel time corresponding to the demand node; This is the time sensitivity coefficient; The cluster demand of the demand node corresponding to the superedge; The average traffic congestion index of the super-edge covered road nodes is obtained through the historical traffic flow data; This is the demand weighting coefficient. It is the traffic weighting coefficient, and and The sum of is 1; Calculate the dynamic weight value of each hyperedge; the higher the dynamic weight value, the higher the priority of the corresponding spatiotemporal demand cluster, the better the passage conditions, and the more likely it is to be covered in the route planning. The dynamic weight values ​​are assigned to the corresponding hyperedges to construct the weighted dynamic hypergraph structure.

[0013] According to one embodiment of the present invention, based on the weighted dynamic hypergraph structure, a two-layer graph convolutional network is constructed and inter-layer convolution operations are performed by inputting the hypergraph adjacency matrix and node attribute matrix to extract the spatial association, demand intensity, and environmental constraint features of nodes, generating a hypergraph node embedding vector set. Based on the hypergraph node embedding vector set, a meta-learning network containing two fully connected layers and a normalized exponential function layer is constructed to complete the nonlinear mapping and normalization processing from node features to multi-objective weights, outputting dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights. The steps of achieving adaptive dynamic allocation of multi-objective optimization weights include: Extract the numbering order of all nodes and the attribute dimensions of the hypergraph nodes in the weighted dynamic hypergraph structure, and construct a two-layer graph convolutional network; set the input layer dimension of the graph convolutional network to be completely matched with the attribute dimensions of the hypergraph nodes, set the first hidden layer dimension of the graph convolutional network to 64 dimensions, the second output layer dimension to 32 dimensions, and set the activation function of both layers of the graph convolutional network to be a linear rectified function, thereby completing the initial configuration of the two-layer graph convolutional network; Based on the initial hypergraph topology and the weighted dynamic hypergraph structure, the association relationships of all nodes and the dynamic weights of the corresponding hyperedges are statistically analyzed to generate a hypergraph adjacency matrix. Based on the customized bus route planning dataset, all nodes' full-dimensional standardized attributes are extracted and arranged in the same order as the hypergraph adjacency matrix to generate a node attribute matrix. The hypergraph adjacency matrix and the node attribute matrix are input into the initialized two-layer graph convolutional network. Inter-layer convolution is performed in the first graph convolutional layer to fuse the association information between the node's own attributes and its first-order neighboring nodes, generating a 64-dimensional intermediate feature vector. This intermediate feature vector is then input into the second graph convolutional layer to perform inter-layer convolution, fusing the global association information between second-order neighboring nodes, thereby generating a corresponding 32-dimensional node embedding vector for each node. ; All node embedding vectors are arranged in the same numbering order as the hypergraph adjacency matrix and the node attribute matrix. The nodes are arranged in an ordered manner to obtain the set of embedding vectors for the hypergraph nodes; The meta-learning network is constructed, and the core weight calculation formula of the meta-learning network is as follows: ; in, It is a multilayer perceptron used to implement nonlinear mapping between node features and weights; This is a normalized exponential function used to probabilistically assign initial weights; The set of embedding vectors for the hypergraph nodes is input into the meta-learning network. First, the node features are nonlinearly mapped through two fully connected layers, transforming the high-dimensional node features into preliminary multi-objective weights. Then, the normalized exponential function layer performs normalization processing, outputting the dynamic weight coefficients optimized for multi-objective goals. These dynamic weight coefficients include path cost weights. Vehicle fixed cost weight And passenger satisfaction penalty weighting And the sum of the three is 1.

[0014] According to one embodiment of the present invention, the overall objective function of the multi-objective optimization is: ; in, The set of feasible paths For demand clusters To demand cluster The aforementioned road segment toll cost data, For decision variables, representing whether to choose arrive The path; For vehicle types, For the first The number of vehicles of this type deployed, The first of the bus vehicle operation parameters Fixed cost per vehicle for this type of vehicle; For a set of demand clusters, For the first The actual number of passengers served by each demand cluster The rated passenger capacity of a single vehicle; , , The dynamic weighting coefficient; The hard constraints include: demand coverage constraints, vehicle capacity constraints, total transport capacity constraints, route length constraints, and time window constraints.

[0015] Secondly, embodiments of this application also provide a customized intelligent bus route planning system based on route feedback collection, the system comprising: Acquisition module: used to acquire customized public transport travel demand data and matching multimodal environmental data to build an original spatiotemporal demand database; Data cleaning module: used to perform outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database to obtain a cleaned spatiotemporal demand dataset; Data standardization module: used to standardize the cleaned spatiotemporal demand dataset, and to construct a standardized spatiotemporal object for boarding point and a standardized spatiotemporal object for disembarkation point with a common unique demand identifier for each valid demand data, forming a set of standardized spatiotemporal objects for boarding point and a set of standardized spatiotemporal objects for disembarkation point respectively; Clustering parameter adaptation and similarity calculation module: used to adaptively match and set clustering core parameters according to the regional function, travel time and data scale of the standardized spatiotemporal objects of the boarding points; and to calculate the comprehensive similarity of any two standardized spatiotemporal objects of the boarding points based on the set of standardized spatiotemporal objects of the boarding points and the clustering core parameters, and to construct a comprehensive similarity matrix of the boarding points. Spatiotemporal clustering module: used to calculate the set of spatiotemporal demand clusters for valid boarding points and the set of spatiotemporal demand clusters for valid alighting points; Attribute encapsulation module: used to complete full-dimensional attribute encapsulation for all the valid boarding point spatiotemporal demand clusters and the valid alighting point spatiotemporal demand clusters respectively, to obtain fully encapsulated boarding point spatiotemporal demand clusters and fully encapsulated alighting point spatiotemporal demand clusters, thereby forming a fully encapsulated boarding point spatiotemporal demand cluster set and a fully encapsulated alighting point spatiotemporal demand cluster set. Matrix construction module: used to construct the boarding point cluster-alighting point cluster OD demand matrix, and output the set of spatiotemporal demand clusters of the fully enclosed boarding point, the set of spatiotemporal demand clusters of the fully enclosed alighting point, and the OD demand matrix of the boarding point cluster-alighting point cluster that can be used for route planning; The data processing module is used to construct a basic dataset for customized bus route planning based on the set of spatiotemporal demand clusters of fully encapsulated boarding points, the set of spatiotemporal demand clusters of fully encapsulated alighting points, the OD demand matrix of the boarding point cluster and alighting point cluster, and the basic data supporting route planning. The Hypergraph Topology module is used to define demand nodes, road nodes, environmental event nodes, and hyperedge mapping rules, construct the initial hypergraph topology structure, and realize cross-dimensional unified association of demand, road network, and environment. The hypergraph weight calculation module is used to construct a dynamic weight function for each hyperedge and complete the weight assignment, generating a weighted dynamic hypergraph structure to realize the quantitative representation of the priority and passage constraints of different requirement clusters. The hypergraph feature extraction module is used to construct a two-layer graph convolutional network and input the hypergraph adjacency matrix and node attribute matrix to perform inter-layer convolution operations to extract the spatial association, demand intensity and environmental constraint features of nodes and generate a set of hypergraph node embedding vectors. The weight allocation module is used to construct a meta-learning network containing two fully connected layers and a normalized exponential function layer. It completes the nonlinear mapping and normalization of node features to multi-objective weights and outputs dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights, thereby realizing the adaptive dynamic allocation of multi-objective optimization weights. The multi-objective mixed-integer programming module is used to construct the overall objective function for multi-objective optimization and inject hard constraints to form a complete multi-objective mixed-integer programming model. The scheme generation module is used to solve the multi-objective mixed integer programming model using an integer programming solver to obtain multiple initial route schemes, calculate the core operation indicators of each initial route scheme and complete compliance verification, and output a customized bus candidate route scheme pool after eliminating invalid schemes.

[0016] Thirdly, embodiments of this application also provide a computer device, the computer device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform the steps of the method described above when executing the instructions.

[0017] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method described above.

[0018] The technical solution provided in this application has the following advantages: 1. The customized bus intelligent route planning method based on route opinion collection proposed in this application integrates spatial location and travel time dimension data to complete clustering, which can capture the fluctuation of travel demand at different times and make the clustering results fit the actual travel patterns of users.

[0019] 2. The customized bus intelligent route planning method based on route opinion collection proposed in this application can adaptively match the core parameters of clustering according to the function of the area where the demand point is located, adapt to the demand characteristics of different urban and rural scenarios, and avoid the problems of insufficient clustering in urban areas and excessive segmentation in rural areas.

[0020] 3. The customized bus intelligent route planning method based on route opinion collection proposed in this application integrates multimodal environmental data such as road network, weather, traffic conditions, and special events during the data cleaning stage. This completes the removal of outliers, deduplication of duplicate data, and correction of demand time windows, reducing the interference of invalid or erroneous data on clustering accuracy and ensuring that subsequent route planning meets actual travel needs.

[0021] 4. The customized bus intelligent route planning method based on route opinion collection proposed in this application can adaptively match and set the core clustering parameters according to the regional functions, travel time and data scale of the demand points during the clustering process. At the same time, it combines spatial correlation weight and temporal correlation weight to calculate the comprehensive similarity between demand points and construct a corresponding similarity matrix to complete the clustering. This achieves spatiotemporal dual-dimensional fusion clustering, adapts to the travel demand characteristics of different scenarios and time periods, and makes the clustering results fit the real travel patterns of users.

[0022] 5. The customized bus intelligent route planning method based on route opinion collection in this application filters spatiotemporal core objects by traversing and judging during the clustering process, and generates demand clusters iteratively based on the core objects. At the same time, it performs secondary verification and filtering on potential noise points in the first round of judgment, which can reduce the situation where effective travel demand is misjudged as isolated noise points. While ensuring the clustering effect, it improves the utilization rate of effective demand data.

[0023] 6. The intelligent route planning method for customized buses based on route opinion collection proposed in this application encapsulates the full-dimensional attributes of the demand clusters after clustering, and constructs an OD demand matrix based on the unique demand identifier. The output clustering results can be used for customized bus route planning, which improves the practicality and feasibility of the solution.

[0024] 7. The intelligent route planning method for customized buses based on route opinion collection proposed in this application adopts a multi-objective collaborative optimization architecture, which takes into account route cost, vehicle fixed cost and passenger satisfaction at the same time. It avoids the problem of deterioration of operation indicators caused by single-objective optimization, and the generated route plan is more in line with the needs of commercial operation of customized buses.

[0025] 8. The customized bus intelligent route planning method based on route opinion collection in this application realizes the unified correlation modeling of demand, road network and environment multi-dimensional elements through hypergraph dynamic modeling, which breaks through the limitation of traditional simple graph pairwise correlation. It can integrate dynamic constraints such as historical traffic and severe weather into the route planning process, and the generated route plan is more in line with the actual road traffic conditions.

[0026] 9. The customized bus intelligent route planning method based on route opinion collection in this application realizes the dynamic allocation of multi-objective optimization weights through meta-learning network, which replaces the traditional manual experience parameter tuning method. It can adjust the priority according to the demand intensity and traffic conditions of different scenarios, and has stronger adaptability in large-scale complex road networks, improving the convergence speed and solution quality of the model.

[0027] 10. The customized bus intelligent route planning method based on route opinion collection in this application realizes the quantitative representation of the priority and traffic constraints of different spatiotemporal demand clusters through a weighted dynamic hypergraph structure. It can reflect the time urgency, demand and traffic difficulty of different spatiotemporal demand clusters, ensuring that areas with high timeliness, high demand and good traffic conditions are prioritized to be covered by routes, thereby improving the allocation efficiency of public transportation resources.

[0028] 11. The customized intelligent bus route planning method based on route opinion collection in this application obtains a set of hypergraph node embedding vectors through inter-layer convolution operation, which automatically captures implicit relationships in large-scale complex road networks that are difficult to define manually, and significantly improves the comprehensiveness and accuracy of feature expression.

[0029] 12. The customized bus intelligent route planning method based on route opinion collection in this application outputs multiple candidate schemes instead of a single optimal scheme, which enables operators to flexibly select the most suitable route scheme according to the operational goals, temporary adjustment needs or resource allocation at different times, significantly improving the system's adaptability in dynamic traffic environments.

[0030] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is the overall flowchart of the intelligent route planning method for customized buses based on route feedback collected in this application.

[0033] Figure 2 This is an architecture diagram of the customized intelligent bus route planning system based on route feedback collected in this application.

[0034] Figure 3 This is a schematic diagram of the user terminal interface for this application.

[0035] Figure 4 This is another screenshot of the user terminal interface for this application. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0037] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0038] The present application will now be described in detail with reference to the accompanying drawings and embodiments.

[0039] Example like Figure 1 As shown, this embodiment provides a customized intelligent bus route planning method based on route feedback collection, including the following steps: S101: Obtain customized public transport travel demand data and matching multimodal environmental data to construct an original spatiotemporal demand database; S102: Based on the multimodal environment data, perform outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database to obtain a cleaned spatiotemporal demand dataset. S103: Standardize the cleaned spatiotemporal demand dataset, and construct a standardized spatiotemporal object for the boarding point and a standardized spatiotemporal object for the alighting point with a common unique demand identifier for each valid demand data, forming a set of standardized spatiotemporal objects for the boarding point and a set of standardized spatiotemporal objects for the alighting point respectively. S104: Based on the regional function, travel time period and data scale of the standardized spatiotemporal object of the boarding point, adaptively match and set the core parameters of clustering; Based on the standardized spatiotemporal object set of the boarding point and the clustering core parameters, calculate the comprehensive similarity between any two standardized spatiotemporal objects of the boarding point and construct a comprehensive similarity matrix of the boarding point. S105: Based on the calibrated boarding point comprehensive similarity matrix and the minimum number of cluster points in the clustering core parameters, traverse, judge and filter the spatiotemporal core objects and potential noise points of the boarding points; iteratively generate the spatiotemporal demand clusters of the boarding points based on the spatiotemporal core objects of the boarding points, and after completing the secondary verification and filtering of the potential noise points, obtain the set of effective spatiotemporal demand clusters of the boarding points. S106: Based on the standardized spatiotemporal object set of the drop-off point and the clustering core parameters, replace the execution subject in S104 to S105 with the standardized spatiotemporal object of the drop-off point and execute again to obtain a set of effective spatiotemporal demand clusters of drop-off points. S107: Complete full-dimensional attribute encapsulation for all the valid boarding point spatiotemporal demand clusters and the valid alighting point spatiotemporal demand clusters respectively, to obtain fully encapsulated boarding point spatiotemporal demand clusters and fully encapsulated alighting point spatiotemporal demand clusters, thereby forming a fully encapsulated boarding point spatiotemporal demand cluster set and a fully encapsulated alighting point spatiotemporal demand cluster set. Both the fully enclosed boarding point spatiotemporal demand cluster set and the fully enclosed disembarkation point spatiotemporal demand cluster set include: spatial attributes, temporal attributes, demand intensity attributes, and association attributes; wherein, the spatial attributes include: cluster center coordinates, geographical coverage boundary, and a list of associated road nodes; the temporal attributes include: a set of timestamps and time distribution characteristics; the demand intensity attributes include: cluster demand quantity, demand density, and priority weight; and the association attributes include original demand data reference links and environmental event tags; S108: Based on the unique demand identifier, the set of spatiotemporal demand clusters of fully enclosed boarding points and the set of spatiotemporal demand clusters of fully enclosed alighting points, construct the boarding point cluster-alighting point cluster OD demand matrix, and output the set of spatiotemporal demand clusters of fully enclosed boarding points, the set of spatiotemporal demand clusters of fully enclosed alighting points and the set of boarding point cluster-alighting point cluster OD demand matrix that can be used for route planning. S109: Based on the fully enclosed boarding point spatiotemporal demand cluster set, the fully enclosed alighting point spatiotemporal demand cluster set, the boarding point cluster-alighting point cluster OD demand matrix, and the supporting basic data for route planning, construct a basic dataset for customized bus route planning; wherein, the fully enclosed boarding point spatiotemporal demand cluster set, the fully enclosed alighting point spatiotemporal demand cluster set, and the boarding point cluster-alighting point cluster OD demand matrix are merged into a spatiotemporal demand cluster set; the supporting basic data for route planning includes urban road network topology data, road construction and road closure event data, severe weather data, road segment toll cost data, historical traffic flow data, and bus vehicle operation parameters associated with the spatiotemporal demand clusters. S110: Based on the customized bus route planning basic dataset, define demand nodes, road nodes, environmental event nodes and hyperedge mapping rules, construct an initial hypergraph topology, and realize cross-dimensional unified association of demand, road network and environment; S111: By constructing a dynamic weight function for each hyperedge and assigning weights, a weighted dynamic hypergraph structure is generated, enabling a quantitative representation of the priority and passage constraints of different demand clusters. S112: Based on the weighted dynamic hypergraph structure, a two-layer graph convolutional network is constructed and inter-layer convolution is performed by inputting the hypergraph adjacency matrix and the node attribute matrix to extract the spatial association, demand intensity and environmental constraint features of the nodes, and generate a set of hypergraph node embedding vectors. S113: Based on the set of embedded vectors of the hypergraph nodes, a meta-learning network containing two fully connected layers and a normalized exponential function layer is constructed to complete the nonlinear mapping and normalization of node features to multi-objective weights, and output dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights, thereby realizing the adaptive dynamic allocation of multi-objective optimization weights. S114: Based on the customized bus route planning basic dataset, the weighted dynamic hypergraph structure, and the dynamic weight coefficients, construct a multi-objective optimization overall objective function and inject hard constraints to form a complete multi-objective mixed integer programming model; S115: Use an integer programming solver to solve the multi-objective mixed integer programming model to obtain multiple initial route schemes. Calculate the core operation indicators of each initial route scheme and complete compliance verification. After eliminating invalid schemes, output a pool of customized bus candidate route schemes.

[0040] The following details each step of the intelligent route planning method for customized buses based on route feedback collection: S101: Obtain customized public transport travel demand data and matching multimodal environmental data to construct an original spatiotemporal demand database; In one possible implementation, step S101 includes the following steps: Step 1011: Through the user terminal (e.g., Figure 3 and Figure 4 (As shown) Obtain the customized bus travel demand data. Each piece of customized bus travel demand data includes the unique demand identifier, the latitude and longitude coordinates of the boarding point, the latitude and longitude coordinates of the alighting point, the user's expected boarding time window, and the user's expected alighting time window. Step 1012: Obtain the multimodal environment data that matches the customized public transportation travel demand data. The multimodal environment data includes urban road network topology data, weather warning data, historical traffic status data, and special event identifier data for holidays and the start of the school season. Step 1013: Perform temporal and spatial correlation matching between all the acquired customized bus travel demand data and the multimodal environment data. Using the user's expected boarding time window and the user's expected alighting time window as the time reference, and the boarding point latitude and longitude coordinates and the alighting point latitude and longitude coordinates as the spatial reference, bind the corresponding spatiotemporal range of the multimodal environment data to each customized bus travel demand data, and construct the original spatiotemporal demand database.

[0041] The customized bus intelligent route planning method based on route opinion collection proposed in this application integrates spatial location and travel time dimension data to complete clustering, which can capture the fluctuation of travel demand at different times and make the clustering results fit the actual travel patterns of users.

[0042] S102: Based on the multimodal environment data, perform outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database to obtain a cleaned spatiotemporal demand dataset. In one possible implementation, step S102 includes the following steps: Step 1021: Perform outlier removal operation: Based on the urban road network topology data, calculate the vertical distance from the latitude and longitude coordinates of each boarding point and the latitude and longitude coordinates of each alighting point to the nearest road network, and remove data with a vertical distance exceeding 500 meters. Step 1022: Perform duplicate request data deduplication: For multiple identical request data submitted by the same user, only one valid data is retained, and all other duplicate data is removed; Step 1023: Perform demand time window correction operation: Based on the weather warning data, if there is severe weather in the corresponding spatiotemporal range, the start time of the user's expected boarding time window is extended forward by 10% and the end time of the user's expected alighting time window is extended backward by 10%; based on the historical traffic status data, if the corresponding spatiotemporal range is a normally congested road section, the start time of the user's expected boarding time window is extended forward by 15% and the end time of the user's expected alighting time window is extended backward by 15%; based on the holiday and school opening season special event identifier data, if the corresponding spatiotemporal range is in the holiday travel peak or school opening season, the start time of the user's expected boarding time window is extended forward by 20% and the end time of the user's expected alighting time window is extended backward by 20%; if a customized bus travel demand data simultaneously meets multiple correction conditions, the largest extension time among the conditions is taken, and the extension time is not superimposed; the upper limit threshold of the extension time is 30 minutes.

[0043] The customized bus intelligent route planning method based on route opinion collection proposed in this application integrates multimodal environmental data such as road network, weather, traffic conditions, and special events during the data cleaning stage. This completes the removal of outliers, deduplication of duplicate data, and correction of demand time windows, reducing the interference of invalid or erroneous data on clustering accuracy and ensuring that subsequent route planning meets actual travel needs.

[0044] S103: Standardize the cleaned spatiotemporal demand dataset, and construct a standardized spatiotemporal object for the boarding point and a standardized spatiotemporal object for the alighting point with a common unique demand identifier for each valid demand data, forming a set of standardized spatiotemporal objects for the boarding point and a set of standardized spatiotemporal objects for the alighting point respectively. In one possible implementation, the structure of the standardized spatiotemporal object of the boarding point is as follows: ; in, This is the unique requirement identifier. Here are the longitude coordinates of the boarding point. Here are the latitude coordinates of the boarding point. This is the midpoint timestamp of the user's expected boarding time window after the demand time window has been corrected. The structure of the standardized spatiotemporal object of the drop-off point is as follows: ; in, The unique demand identifier is shared with the standardized spatiotemporal object of the boarding point. Here are the longitude coordinates of the drop-off point. Here are the latitude coordinates of the drop-off point. This is the midpoint timestamp of the user's expected drop-off time window after the demand time window has been corrected.

[0045] S104: Based on the regional function, travel time and data scale of the standardized spatiotemporal objects of the boarding points, adaptively match and set the core parameters of clustering; based on the set of standardized spatiotemporal objects of the boarding points and the core parameters of clustering, calculate the comprehensive similarity of any two standardized spatiotemporal objects of the boarding points, and construct a comprehensive similarity matrix of the boarding points. In one possible implementation, the clustering core parameter includes the spatial decay radius. Time decay coefficient and minimum number of cluster points The steps for adaptive matching settings of the clustering core parameters include: Based on the functional areas of the standardized spatiotemporal objects at the boarding points, they are classified into commercial and industrial parks, residential areas, and rural areas, and corresponding spatial attenuation radii are set for each. ; Specifically, for spatiotemporal object points in commercial and industrial parks, set Within a 300-meter radius, demand is dense in the city's core area, and users have short walking distances, requiring precise aggregation; for spatiotemporal target points in residential areas, [the following parameters are set]. The maximum distance is 500 meters, which is slightly longer for residential users to walk, and is an aggregation range suitable for commuting needs; for spatiotemporal target points in rural areas, a specific setting is made. The distance is 1000 meters. Rural demand is scattered, and a larger aggregation range is needed to ensure the feasibility of line operation.

[0046] Here, this application can adaptively match the core clustering parameters according to the function of the area where the demand point is located, adapt to the demand characteristics of different urban and rural scenarios, and avoid the problems of insufficient clustering in urban areas and excessive segmentation in rural areas.

[0047] Based on the standardized spatiotemporal objects of the boarding points, the travel time periods are classified into weekday morning and evening peak hours, weekday off-peak hours, and holidays, and corresponding time decay coefficients are set for each. ; Specifically, for spatiotemporal target points during weekday morning and evening rush hours, set For 15-minute commutes with rigid demand, users have extremely high requirements for consistent boarding times; only those with similar boarding times can be grouped together on the same bus. For specific time and location points during weekday off-peak hours, [the following is defined / set / implied]. The travel time is set at 30 minutes, allowing for more flexibility during off-peak hours; however, for holidays, a more flexible timeframe is set. The time limit is 60 minutes, and there are no strict time requirements for leisure travel, allowing for a greater aggregation of time differences.

[0048] Based on the data scale of the standardized spatiotemporal object of the boarding point, the minimum number of cluster points is calculated. : ; in, The total number of standardized spatiotemporal objects for the boarding point within the standardized spatiotemporal object set for the boarding point. This represents the total number of pre-set customized bus routes. Based on the standardized spatiotemporal object set of boarding points and the clustering core parameters, the steps for calculating the comprehensive similarity between any two standardized spatiotemporal objects of boarding points and constructing the comprehensive similarity matrix of boarding points include: Iterate through all pairwise combinations of the standardized spatiotemporal objects of the boarding points within the set of standardized spatiotemporal objects of the boarding points. and Calculate the spatial correlation weight between the two points respectively. Time-related weights ; Among them, the spatial correlation weight : ; in, Standardize the spatiotemporal object of the boarding point and The Euclidean distance is calculated using latitude and longitude coordinates; The spatial attenuation radius; The time-related weight : ; in, The absolute difference between the timestamps of the two points; The time decay coefficient is mentioned above; Based on the spatial correlation weight Time-related weights Calculate the comprehensive similarity : ; The overall similarity The range of values ​​is A higher value indicates a stronger spatial proximity and temporal consistency between the two points, and a higher cluster correlation. Complete the comprehensive similarity of all pairwise combinations of spatiotemporal objects within the standardized spatiotemporal object set of the boarding point. After calculation, a comprehensive similarity matrix for the boarding points is constructed. This matrix is ​​a two-dimensional square matrix with equal rows and columns. Each row and column corresponds to a standardized spatiotemporal object of the boarding point, and each cell element represents the comprehensive similarity score. .

[0049] The customized bus intelligent route planning method based on route opinion collection proposed in this application can adaptively match and set the core clustering parameters according to the regional functions, travel time and data scale of the demand points during the clustering process. At the same time, it combines spatial correlation weight and temporal correlation weight to calculate the comprehensive similarity between demand points and construct a corresponding similarity matrix to complete the clustering. This achieves spatiotemporal dual-dimensional fusion clustering, adapts to the travel demand characteristics of different scenarios and time periods, and makes the clustering results fit the real travel patterns of users.

[0050] S105: Based on the calibrated boarding point comprehensive similarity matrix and the minimum number of cluster points in the clustering core parameters, traverse, judge and filter the spatiotemporal core objects and potential noise points of the boarding points; iteratively generate the spatiotemporal demand clusters of the boarding points based on the spatiotemporal core objects of the boarding points, and after completing the secondary verification and filtering of the potential noise points, obtain the set of effective spatiotemporal demand clusters of the boarding points. In one possible implementation, step S105 includes the following steps: S1051: Initialize the classification state of all the standardized spatiotemporal objects at the boarding points to unclassified and the processing state to unprocessed; initialize the initial value of the auto-incrementing cluster number sequence to 0; and initialize the set of potential noise points to empty. The classification status includes: unclassified, classified, and marked; where unclassified means that the object has not completed the verification and attribution determination; classified means that the object has been assigned to the determined spatiotemporal demand cluster of the boarding point; marked means that the object has completed the first round of verification and has been identified as a potential noise point, and is waiting for the second verification and filtering. S1052: Traverse each of the standardized spatiotemporal objects of the boarding point in the set of standardized spatiotemporal objects of the boarding point and check its classification status: if it is classified, continue to traverse the next standardized spatiotemporal object of the boarding point; if it is unclassified, treat it as an unclassified spatiotemporal object Pi and execute S1053. S1053: For the unclassified spatiotemporal object Pi, select the comprehensive similarity scores with the unclassified spatiotemporal object Pi from the comprehensive similarity matrix of the boarding point. All standardized spatiotemporal objects of the boarding points that are greater than or equal to the neighborhood similarity threshold constitute the spatiotemporal neighborhood point set of the unclassified spatiotemporal object Pi; the neighborhood similarity threshold is a preset configurable parameter; S1054: Count the total number of standardized spatiotemporal objects of the boarding point within the spatiotemporal neighborhood point set. If the total number is less than the minimum clustering point number... If the unclassified spatiotemporal object Pi is determined to be a potential noise point, it is added to the potential noise point set, and its classification status is updated to marked. Then, return to S1052 to continue the traversal operation. If the total number is greater than or equal to the minimum number of cluster points If so, the unclassified spatiotemporal object Pi is determined to be the core spatiotemporal object of the boarding point, the auto-incrementing cluster number is assigned to it, and the process proceeds to S1055. S1055: All the core spatiotemporal objects of the boarding point, as well as all the standardized spatiotemporal objects of the boarding point in its spatiotemporal neighborhood point set, are classified into the spatiotemporal demand cluster of the boarding point corresponding to the auto-incrementing cluster number, and their classification status is uniformly updated to classified. S1056: Iterate through all standardized spatiotemporal objects of the current boarding point spatiotemporal demand cluster whose processing status is unprocessed. For each currently traversed cluster object Px, filter out the comprehensive similarity between the cluster object Px and the current cluster object Px from the comprehensive similarity matrix of the boarding points. All standardized spatiotemporal objects of the boarding points that are greater than or equal to the neighborhood similarity threshold constitute the spatiotemporal neighborhood point set of the intra-cluster object Px; Count the total number of objects in the spatiotemporal neighborhood point set of the cluster object Px. If the total number is greater than or equal to the minimum clustering point number... If the object Px in the cluster is determined to be the new spatiotemporal core object of the boarding point, all objects in the spatiotemporal neighborhood point set whose classification status is unclassified are simultaneously assigned to the spatiotemporal demand cluster of the boarding point corresponding to the current auto-incrementing cluster number, and their classification status is updated to classified. After completing the above judgment and attribution operations, the processing status of the current cluster object Px is updated to processed; the above traversal operation is continued until the processing status of all objects in the boarding point spatiotemporal demand cluster is processed, thereby completing the generation of the boarding point spatiotemporal demand cluster, and the value of the auto-incrementing cluster number is incremented by one, and then proceeding to S1057. S1057: If there are still unclassified standardized spatiotemporal objects in the set of standardized spatiotemporal objects of the boarding point, return to S1052 to continue the traversal operation; if all the standardized spatiotemporal objects of the boarding point are classified or marked, proceed to S1058. S1058: For all potential noise points in the set of potential noise points, perform secondary verification and filtering: calculate the comprehensive similarity between the potential noise point and all generated spatiotemporal core objects of the boarding point. Take the maximum value among them; If the maximum value is greater than or equal to the noise point attribution threshold, then the potential noise point is assigned to the spatiotemporal demand cluster of the boarding point spatiotemporal core object corresponding to the maximum value, and its classification status is updated to classified; the noise point attribution threshold is a preset configurable parameter and is less than the neighborhood similarity threshold. If the maximum value is less than the noise point attribution threshold, the potential noise point is marked as a spatiotemporally isolated noise point and is not included in any of the boarding point spatiotemporal demand clusters. S1059: After completing the generation of all the aforementioned spatiotemporal demand clusters for boarding points and the filtering of noise points, the set of effective spatiotemporal demand clusters for boarding points is obtained.

[0051] Optionally, the neighborhood similarity threshold is 0.3, and the noise point attribution threshold is 0.2.

[0052] The customized bus intelligent route planning method based on route opinion collection proposed in this application filters spatiotemporal core objects through traversal judgment during the clustering process, and generates demand clusters iteratively based on the core objects. At the same time, it performs secondary verification and filtering on potential noise points identified in the first round of judgment, which can reduce the situation where effective travel demand is misjudged as isolated noise points. While ensuring the clustering effect, it improves the utilization rate of effective demand data.

[0053] S106: Based on the standardized spatiotemporal object set of the drop-off point and the clustering core parameters, replace the execution subject in S104 to S105 with the standardized spatiotemporal object of the drop-off point and execute again to obtain a set of effective spatiotemporal demand clusters of drop-off points. S107: Perform full-dimensional attribute encapsulation for all the valid boarding point spatiotemporal demand clusters and the valid alighting point spatiotemporal demand clusters respectively, to obtain fully encapsulated boarding point spatiotemporal demand clusters and fully encapsulated alighting point spatiotemporal demand clusters, thereby forming a fully encapsulated boarding point spatiotemporal demand cluster set and a fully encapsulated alighting point spatiotemporal demand cluster set; both the fully encapsulated boarding point spatiotemporal demand cluster set and the fully encapsulated alighting point spatiotemporal demand cluster set include: spatial attributes, temporal attributes, demand intensity attributes, and correlation attributes. Specifically, the spatial attributes include: cluster center coordinates, geographic coverage boundary, and a list of associated road nodes. The cluster center coordinates are the average of the latitude and longitude coordinates of all standardized spatiotemporal objects within the cluster. The geographic coverage boundary is calculated by first determining the Euclidean distance from each standardized spatiotemporal object within the cluster to the cluster center coordinates, then taking the maximum value as the coverage radius. A circular area with the cluster center coordinates as the center and the coverage radius as the radius is the geographic coverage boundary of the cluster. Alternatively, the latitude and longitude coordinates of all standardized spatiotemporal objects within the cluster can be connected into a minimum closed polygon using a convex hull algorithm, serving as the precise geographic coverage boundary of the cluster. The list of associated road nodes is formed by matching the original collected urban road network topology data with the cluster center coordinates and geographic coverage boundary, extracting all accessible urban road intersections and road segments within the geographic coverage boundary, and deduplicating them to create the list of associated road nodes for the cluster.

[0054] The time attributes include: a timestamp set and time distribution characteristics. The timestamp set is formed by directly extracting the timestamps of all standardized spatiotemporal objects within the corresponding cluster, removing duplicates, and sorting them chronologically. The timestamps for the boarding point spatiotemporal demand cluster are the midpoint timestamps of the corrected user's expected boarding time window, and the timestamps for the alighting point spatiotemporal demand cluster are the midpoint timestamps of the corrected user's expected alighting time window. The time distribution characteristics are calculated based on the timestamp set and include three core components: first, the timestamp mean, which is the average of all timestamps in the set, representing the core travel time of users in the cluster; second, the timestamp variance, used to measure the dispersion of user travel time within the cluster, with smaller variance indicating more concentrated user travel time; and third, the peak travel time window, which involves frequency statistics on the timestamp set, selecting consecutive time periods where the number of users exceeds 60% of the total number of people in the cluster as the peak travel time window for that cluster.

[0055] The demand intensity attributes include: cluster demand, demand density, and priority weight. Cluster demand directly represents the total number of standardized spatiotemporal objects within a cluster, corresponding to the total number of passengers that the virtual station can serve. Demand density is calculated by dividing the cluster demand by the area of ​​the cluster's geographical coverage boundary, yielding the demand per unit geographical area, which measures the density of demand in the region. Priority weight is calculated based on cluster demand and peak travel time windows, with the cluster demand as the base weight. For example, if the peak travel time window falls during weekday morning and evening rush hours, the weight is multiplied by a coefficient of 1.5; a higher weight indicates a higher priority for the cluster's route coverage.

[0056] The associated attributes include original demand data reference links and environmental event tags. Original demand data reference links: Each standardized spatiotemporal object within the cluster corresponds to a single original customized public transport travel demand data entry through a unique demand identifier. Storage reference links for the corresponding original demand data are bound to all objects within the cluster. These links are then aggregated to form a set of original demand data reference links for the cluster, allowing for the tracing of the complete original information for each demand. Environmental event tags: Multimodal environmental data corresponding to all standardized spatiotemporal objects within the cluster are extracted and aggregated to form environmental event tags for the cluster. These tags include weather, historical traffic conditions, and special events such as holidays and school opening seasons within the corresponding spatiotemporal range.

[0057] S108: Based on the unique demand identifier, the set of spatiotemporal demand clusters of fully enclosed boarding points and the set of spatiotemporal demand clusters of fully enclosed alighting points, construct the boarding point cluster-alighting point cluster OD demand matrix, and output the set of spatiotemporal demand clusters of fully enclosed boarding points, the set of spatiotemporal demand clusters of fully enclosed alighting points and the set of boarding point cluster-alighting point cluster OD demand matrix that can be used for route planning. In one possible implementation, the step of constructing the boarding point cluster-alighting point cluster OD demand matrix in step S108 includes: Based on the unique demand identifier, establish a mapping relationship between the standardized spatiotemporal object of the boarding point and the standardized spatiotemporal object of the alighting point; The standardized spatiotemporal object of each boarding point is matched with the spatiotemporal demand cluster of the fully encapsulated boarding point to which it belongs, and the standardized spatiotemporal object of each alighting point is matched with the spatiotemporal demand cluster of the fully encapsulated alighting point to which it belongs. A matrix basic framework is constructed using all the spatiotemporal demand clusters of the fully encapsulated loading points as the row dimension and all the spatiotemporal demand clusters of the fully encapsulated unloading points as the column dimension. Based on the mapping relationship and the attribution relationship matching, the total number of effective demands from each of the fully encapsulated boarding point spatiotemporal demand clusters to each of the fully encapsulated disembarking point spatiotemporal demand clusters is counted and filled into the cell; Each row of the matrix is ​​bound to the full-dimensional encapsulation attribute of the corresponding fully encapsulated boarding point spatiotemporal demand cluster, and each column of the matrix is ​​bound to the full-dimensional encapsulation attribute of the corresponding fully encapsulated disembarkation point spatiotemporal demand cluster, thereby forming the boarding point cluster-disembarkation point cluster OD demand matrix.

[0058] The intelligent route planning method for customized buses based on route opinion collection proposed in this application encapsulates the full-dimensional attributes of the demand clusters after clustering, and constructs an OD demand matrix based on unique demand identifiers for boarding point clusters and alighting point clusters. The output clustering results can be used for customized bus route planning, improving the practicality and feasibility of the solution.

[0059] S109: Based on the fully enclosed boarding point spatiotemporal demand cluster set, the fully enclosed alighting point spatiotemporal demand cluster set, the boarding point cluster-alighting point cluster OD demand matrix, and the supporting basic data for route planning, construct a basic dataset for customized bus route planning; wherein, the fully enclosed boarding point spatiotemporal demand cluster set, the fully enclosed alighting point spatiotemporal demand cluster set, and the boarding point cluster-alighting point cluster OD demand matrix are merged into a spatiotemporal demand cluster set; the supporting basic data for route planning includes urban road network topology data, road construction and road closure event data, severe weather data, road segment toll cost data, historical traffic flow data, and bus vehicle operation parameters associated with the spatiotemporal demand clusters. For example, the severe weather data refers to all weather phenomena that have a significant impact on road traffic, specifically including: heavy rain, heavy snow, heavy fog, hail, and other weather that reduces road traffic capacity or increases safety risks.

[0060] S110: Based on the customized bus route planning basic dataset, define demand nodes, road nodes, environmental event nodes and hyperedge mapping rules, construct an initial hypergraph topology, and realize cross-dimensional unified association of demand, road network and environment; In one possible implementation, step S110 includes the following steps: Define the demand node, the road node, and the environmental event node; wherein, the demand node corresponds to the spatiotemporal demand cluster, the road node corresponds to the intersection and road segment endpoint in the urban road network topology data, and the environmental event node corresponds to the road construction and road closure event data and the event node affecting traffic in the severe weather data; Each demand node is mapped to an independent hyperedge, and each hyperedge connects all the road nodes covered by the virtual site corresponding to the demand node, as well as all the environmental event nodes within the time distribution range of the demand node, thereby realizing a cross-dimensional unified association between demand, road network and environment. Based on the definitions of the demand nodes, road nodes, and environmental event nodes, as well as the hyperedge mapping rules, the initial hypergraph topology is constructed, clarifying the relationships between all nodes and the coverage of hyperedges.

[0061] S111: By constructing a dynamic weight function for each hyperedge and assigning weights, a weighted dynamic hypergraph structure is generated, enabling a quantitative representation of the priority and passage constraints of different demand clusters. In one possible implementation, the step includes the following steps: Construct the dynamic weight function, which is as follows: ; in, For the current planning period; The base travel time corresponding to the demand node; This is the time sensitivity coefficient; The cluster demand of the demand node corresponding to the superedge; The average traffic congestion index of the super-edge covered road nodes is obtained through the historical traffic flow data; This is the demand weighting coefficient. It is the traffic weighting coefficient, and and The sum of is 1; Calculate the dynamic weight value of each hyperedge; the higher the dynamic weight value, the higher the priority of the corresponding spatiotemporal demand cluster, the better the passage conditions, and the more likely it is to be covered in the route planning. The dynamic weight values ​​are assigned to the corresponding hyperedges to construct the weighted dynamic hypergraph structure.

[0062] The customized bus intelligent route planning method based on route opinion collection in this application realizes the quantitative representation of the priority and traffic constraints of different spatiotemporal demand clusters through a weighted dynamic hypergraph structure. It can reflect the time urgency, demand and traffic difficulty of different spatiotemporal demand clusters, ensuring that areas with high timeliness, high demand and good traffic conditions are prioritized for route coverage, thereby improving the allocation efficiency of public transportation resources.

[0063] S112: Based on the weighted dynamic hypergraph structure, a two-layer graph convolutional network is constructed and inter-layer convolution is performed by inputting the hypergraph adjacency matrix and the node attribute matrix to extract the spatial association, demand intensity and environmental constraint features of the nodes, and generate a set of hypergraph node embedding vectors. The customized bus intelligent route planning method based on route opinion collection proposed in this application realizes unified correlation modeling of multiple dimensions of demand, road network and environment through hypergraph dynamic modeling. It breaks through the limitation of pairwise correlation of traditional simple graphs and can integrate dynamic constraints such as historical traffic and severe weather into the route planning process, so that the generated route plan is more in line with the actual road traffic conditions.

[0064] In one possible implementation, step S112 includes the following steps: Extract the numbering order of all nodes and the attribute dimensions of the hypergraph nodes in the weighted dynamic hypergraph structure, and construct a two-layer graph convolutional network; set the input layer dimension of the graph convolutional network to be completely matched with the attribute dimensions of the hypergraph nodes, set the first hidden layer dimension of the graph convolutional network to 64 dimensions, the second output layer dimension to 32 dimensions, and set the activation function of both layers of the graph convolutional network to be a linear rectified function, thereby completing the initial configuration of the two-layer graph convolutional network; Based on the initial hypergraph topology and the weighted dynamic hypergraph structure, the association relationships of all nodes and the dynamic weights of the corresponding hyperedges are statistically analyzed to generate a hypergraph adjacency matrix. Based on the customized bus route planning dataset, all nodes' full-dimensional standardized attributes are extracted and arranged in the same order as the hypergraph adjacency matrix to generate a node attribute matrix. The hypergraph adjacency matrix and the node attribute matrix are input into the initialized two-layer graph convolutional network. Inter-layer convolution is performed in the first graph convolutional layer to fuse the association information between the node's own attributes and its first-order neighboring nodes, generating a 64-dimensional intermediate feature vector. This intermediate feature vector is then input into the second graph convolutional layer to perform inter-layer convolution, fusing the global association information between second-order neighboring nodes, thereby generating a corresponding 32-dimensional node embedding vector for each node. ; All node embedding vectors are arranged in the same numbering order as the hypergraph adjacency matrix and the node attribute matrix. The nodes are arranged in an ordered manner to obtain the set of embedding vectors for the hypergraph nodes.

[0065] The customized intelligent bus route planning method based on route opinion collection proposed in this application obtains a set of hypergraph node embedding vectors through inter-layer convolution operations, which automatically captures implicit relationships in large-scale complex road networks that are difficult to define manually, and significantly improves the comprehensiveness and accuracy of feature representation.

[0066] S113: Based on the set of embedded vectors of the hypergraph nodes, a meta-learning network containing two fully connected layers and a normalized exponential function layer is constructed to complete the nonlinear mapping and normalization of node features to multi-objective weights, and output dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights, thereby realizing the adaptive dynamic allocation of multi-objective optimization weights. In one possible implementation, step S113 includes the following steps: The meta-learning network is constructed, and the core weight calculation formula of the meta-learning network is as follows: ; in, It is a multilayer perceptron used to implement nonlinear mapping between node features and weights; This is a normalized exponential function used to probabilistically assign initial weights; The set of embedding vectors of the hypergraph nodes is input into the meta-learning network. The nonlinear mapping of node features is first completed through the two fully connected layers, and the high-dimensional node features are transformed into multi-objective preliminary weights. The normalization process is then performed through the normalized exponential function layer to output the dynamic weight coefficients for multi-objective optimization; the dynamic weight coefficients include path cost weights. Vehicle fixed cost weight And passenger satisfaction penalty weighting And the sum of the three is 1.

[0067] Among them, path cost weight Prioritize the economic efficiency of control paths. The higher the weight, the more the algorithm tends to choose routes with low fuel consumption and short distances; vehicle fixed cost weight. Prioritize controlling vehicle fixed costs. The higher the value, the more the algorithm tends to reduce the total number of vehicles; passenger satisfaction penalty weight. Prioritize passenger satisfaction. The higher the value, the more the algorithm tends to avoid overcrowding and improve passenger comfort.

[0068] When road operating costs change, such as due to rising oil prices, the route cost weight is automatically increased. Prioritize controlling the total operating cost of the route; when the cost of purchasing, leasing, or maintaining buses increases, the weight of fixed costs for vehicles will be automatically increased. Prioritize reducing the total number of vehicles deployed and lowering vehicle operation costs; when operating subsidies increase and overall cost pressure decreases, automatically increase the penalty weight for passenger satisfaction. Prioritize improving service quality; at the same time, avoid extreme weighting of a single objective by using a normalized exponential function to ensure a balanced adaptation of multiple objectives.

[0069] The intelligent route planning method for customized buses based on route feedback collection proposed in this application adopts a multi-objective collaborative optimization architecture, which takes into account route cost, vehicle fixed cost and passenger satisfaction, thus avoiding the deterioration of operating indicators caused by single-objective optimization. The generated route plan is more in line with the needs of commercial operation of customized buses.

[0070] The customized intelligent bus route planning method based on route opinion collection proposed in this application achieves dynamic allocation of multi-objective optimization weights through meta-learning networks, replacing the traditional manual experience-based parameter tuning method. It can adjust priorities according to the demand intensity and traffic conditions of different scenarios, and has stronger adaptability in large-scale complex road networks, improving the convergence speed and solution quality of the model.

[0071] S114: Based on the customized bus route planning basic dataset, the weighted dynamic hypergraph structure, and the dynamic weight coefficients, construct a multi-objective optimization overall objective function and inject hard constraints to form a complete multi-objective mixed integer programming model; In one possible implementation, the overall objective function of the multi-objective optimization is: ; in, The set of feasible paths For demand clusters To demand cluster The aforementioned road segment toll cost data, For decision variables, representing whether to choose arrive The path; For vehicle types, For the first The number of vehicles of this type deployed, The first of the bus vehicle operation parameters Fixed cost per vehicle for this type of vehicle; For a set of demand clusters, For the first The actual number of passengers served by each demand cluster The rated passenger capacity of a single vehicle; , , The dynamic weighting coefficient; The hard constraints include: demand coverage constraints, vehicle capacity constraints, total transport capacity constraints, route length constraints, and time window constraints.

[0072] Specifically, the constraints are as follows: Demand Coverage Constraint: All generated routes must cover at least 80% of high-priority demand clusters, which are demand nodes with a superedge weight greater than 0.5; Vehicle Capacity Constraint: The total number of passengers on a single route must not exceed the total rated passenger capacity of the deployed vehicles, and overloading is prohibited; Total Capacity Constraint: The total number of vehicles deployed on all routes must not exceed the preset maximum number of vehicles; Route Length Constraint: The total mileage of a single route must not exceed the preset maximum route mileage threshold; Time Window Constraint: The total travel time of a single route must match the time distribution window of the corresponding demand cluster and must not exceed the user's expected travel time window.

[0073] S115: Use an integer programming solver to solve the multi-objective mixed integer programming model to obtain multiple initial route schemes. Calculate the core operation indicators of each initial route scheme and complete compliance verification. After eliminating invalid schemes, output a pool of customized bus candidate route schemes.

[0074] For example, the Gurobi integer programming solver is used to solve the model, with the upper limit of the solution iterations set to 1000 times and the solution accuracy threshold set to 0.01. After solving the problem, we obtain the solution set of the model. Each solution corresponds to a complete route plan, which includes the route direction, the stopping order of demand clusters, and the number of vehicles deployed on the route. Calculate the core operational indicators for each initial route plan, including total route mileage, estimated total operating cost, average route load factor, demand coverage, and total travel time. Link the operational indicators to the route plan. Perform compliance verification, eliminate invalid plans, and output a pool of customized bus candidate route plans.

[0075] The customized bus intelligent route planning method based on route opinion collection proposed in this application outputs multiple candidate schemes instead of a single optimal scheme, enabling operators to flexibly select the most suitable route scheme according to different operational goals, temporary adjustment needs or resource allocation, which significantly improves the system's adaptability in dynamic traffic environments.

[0076] like Figure 2 As shown, based on the same application concept, this application embodiment also provides a customized bus intelligent route planning system based on route opinion collection, including an acquisition module: used to acquire customized bus travel demand data and matching multimodal environment data to construct an original spatiotemporal demand database; Data cleaning module: used to perform outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database to obtain a cleaned spatiotemporal demand dataset; Data standardization module: used to standardize the cleaned spatiotemporal demand dataset, and to construct a standardized spatiotemporal object for boarding point and a standardized spatiotemporal object for disembarkation point with a common unique demand identifier for each valid demand data, forming a set of standardized spatiotemporal objects for boarding point and a set of standardized spatiotemporal objects for disembarkation point respectively; Clustering parameter adaptation and similarity calculation module: used to adaptively match and set clustering core parameters according to the regional function, travel time and data scale of the standardized spatiotemporal objects of the boarding points; and to calculate the comprehensive similarity of any two standardized spatiotemporal objects of the boarding points based on the set of standardized spatiotemporal objects of the boarding points and the clustering core parameters, and to construct a comprehensive similarity matrix of the boarding points. Spatiotemporal clustering module: used to calculate the set of spatiotemporal demand clusters for valid boarding points and the set of spatiotemporal demand clusters for valid alighting points; Attribute encapsulation module: used to complete full-dimensional attribute encapsulation for all the valid boarding point spatiotemporal demand clusters and the valid alighting point spatiotemporal demand clusters respectively, to obtain fully encapsulated boarding point spatiotemporal demand clusters and fully encapsulated alighting point spatiotemporal demand clusters, thereby forming a fully encapsulated boarding point spatiotemporal demand cluster set and a fully encapsulated alighting point spatiotemporal demand cluster set. Matrix construction module: used to construct the boarding point cluster-alighting point cluster OD demand matrix, and output the set of spatiotemporal demand clusters of the fully enclosed boarding point clusters, the set of spatiotemporal demand clusters of the fully enclosed alighting point clusters, and the OD demand matrix of the boarding point cluster-alighting point clusters that can be used for route planning; The data processing module is used to construct a basic dataset for customized bus route planning based on the set of spatiotemporal demand clusters of fully encapsulated boarding points, the set of spatiotemporal demand clusters of fully encapsulated alighting points, the OD demand matrix of the boarding point cluster and alighting point cluster, and the basic data supporting route planning. The Hypergraph Topology module is used to define demand nodes, road nodes, environmental event nodes, and hyperedge mapping rules, construct the initial hypergraph topology structure, and realize cross-dimensional unified association of demand, road network, and environment. The hypergraph weight calculation module is used to construct a dynamic weight function for each hyperedge and complete the weight assignment, generating a weighted dynamic hypergraph structure to realize the quantitative representation of the priority and passage constraints of different requirement clusters. The hypergraph feature extraction module is used to construct a two-layer graph convolutional network and input the hypergraph adjacency matrix and node attribute matrix to perform inter-layer convolution operations to extract the spatial association, demand intensity and environmental constraint features of nodes and generate a set of hypergraph node embedding vectors. The weight allocation module is used to construct a meta-learning network containing two fully connected layers and a normalized exponential function layer. It completes the nonlinear mapping and normalization of node features to multi-objective weights and outputs dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights, thereby realizing the adaptive dynamic allocation of multi-objective optimization weights. The multi-objective mixed-integer programming module is used to construct the overall objective function for multi-objective optimization and inject hard constraints to form a complete multi-objective mixed-integer programming model. The scheme generation module is used to solve the multi-objective mixed integer programming model using an integer programming solver to obtain multiple initial route schemes, calculate the core operation indicators of each initial route scheme and complete compliance verification, and output a customized bus candidate route scheme pool after eliminating invalid schemes.

[0077] Based on the same concept, embodiments of this application also provide a computer device, including at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by a processor, the instructions being executed by the at least one processor to cause the at least one processor to implement the above-described method when executing the instructions.

[0078] Based on the same concept, embodiments of this application also provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described method.

[0079] In this embodiment, the computer program, when run by the processor, can also execute other machine-readable instructions to perform other methods as described in the embodiments. For details on the specific execution steps and principles, please refer to the description of the embodiments, which will not be repeated here.

[0080] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0081] 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 units can be selected to achieve the purpose of this embodiment according to actual needs.

[0082] In addition, the functional units in the embodiments provided in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0083] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion 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 this application. 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.

[0084] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0085] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application; and these modifications, changes, 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 this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A customized bus intelligent route planning method based on route feedback collection, characterized in that, include: S101: Obtain customized public transport travel demand data and matching multimodal environmental data to construct an original spatiotemporal demand database; S102: Based on the multimodal environment data, perform outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database to obtain a cleaned spatiotemporal demand dataset. S103: Standardize the cleaned spatiotemporal demand dataset, and construct a standardized spatiotemporal object for the boarding point and a standardized spatiotemporal object for the alighting point with a common unique demand identifier for each valid demand data, forming a set of standardized spatiotemporal objects for the boarding point and a set of standardized spatiotemporal objects for the alighting point respectively. S104: Based on the regional function, travel time period and data scale of the standardized spatiotemporal object of the boarding point, adaptively match and set the core parameters of clustering; Based on the standardized spatiotemporal object set of the boarding point and the clustering core parameters, calculate the comprehensive similarity between any two standardized spatiotemporal objects of the boarding point and construct a comprehensive similarity matrix of the boarding point. S105: Based on the calibrated comprehensive similarity matrix of the boarding points and the minimum number of cluster points in the clustering core parameters, traverse, judge, and filter the spatiotemporal core objects and potential noise points of the boarding points; Based on the core spatiotemporal object of the boarding point, a spatiotemporal demand cluster of boarding points is generated iteratively. After completing the secondary verification and filtering of the potential noise points, a set of effective spatiotemporal demand clusters of boarding points is obtained. S106: Based on the standardized spatiotemporal object set of the drop-off point and the clustering core parameters, replace the execution subject in S104 to S105 with the standardized spatiotemporal object of the drop-off point and execute again to obtain a set of effective spatiotemporal demand clusters of drop-off points. S107: Complete full-dimensional attribute encapsulation for all the valid boarding point spatiotemporal demand clusters and the valid alighting point spatiotemporal demand clusters respectively, to obtain fully encapsulated boarding point spatiotemporal demand clusters and fully encapsulated alighting point spatiotemporal demand clusters, thereby forming a fully encapsulated boarding point spatiotemporal demand cluster set and a fully encapsulated alighting point spatiotemporal demand cluster set. Both the fully enclosed boarding point spatiotemporal demand cluster set and the fully enclosed disembarkation point spatiotemporal demand cluster set include: spatial attributes, temporal attributes, demand intensity attributes, and association attributes; wherein, the spatial attributes include: cluster center coordinates, geographical coverage boundary, and a list of associated road nodes; the temporal attributes include: a set of timestamps and time distribution characteristics; the demand intensity attributes include: cluster demand quantity, demand density, and priority weight; and the association attributes include original demand data reference links and environmental event tags; S108: Based on the unique demand identifier, the set of spatiotemporal demand clusters of fully enclosed boarding points and the set of spatiotemporal demand clusters of fully enclosed alighting points, construct the boarding point cluster-alighting point cluster OD demand matrix, and output the set of spatiotemporal demand clusters of fully enclosed boarding points, the set of spatiotemporal demand clusters of fully enclosed alighting points and the set of boarding point cluster-alighting point cluster OD demand matrix that can be used for route planning. S109: Based on the fully enclosed boarding point spatiotemporal demand cluster set, the fully enclosed alighting point spatiotemporal demand cluster set, the boarding point cluster-alighting point cluster OD demand matrix, and the supporting basic data for route planning, construct a basic dataset for customized bus route planning; wherein, the fully enclosed boarding point spatiotemporal demand cluster set, the fully enclosed alighting point spatiotemporal demand cluster set, and the boarding point cluster-alighting point cluster OD demand matrix are merged into a spatiotemporal demand cluster set; the supporting basic data for route planning includes urban road network topology data, road construction and road closure event data, severe weather data, road segment toll cost data, historical traffic flow data, and bus vehicle operation parameters associated with the spatiotemporal demand clusters. S110: Based on the customized bus route planning basic dataset, define demand nodes, road nodes, environmental event nodes and hyperedge mapping rules, construct an initial hypergraph topology, and realize cross-dimensional unified association of demand, road network and environment; S111: By constructing a dynamic weight function for each hyperedge and assigning weights, a weighted dynamic hypergraph structure is generated, enabling a quantitative representation of the priority and passage constraints of different demand clusters. S112: Based on the weighted dynamic hypergraph structure, a two-layer graph convolutional network is constructed and inter-layer convolution is performed by inputting the hypergraph adjacency matrix and the node attribute matrix to extract the spatial association, demand intensity and environmental constraint features of the nodes, and generate a set of hypergraph node embedding vectors. S113: Based on the set of embedded vectors of the hypergraph nodes, a meta-learning network containing two fully connected layers and a normalized exponential function layer is constructed to complete the nonlinear mapping and normalization of node features to multi-objective weights, and output dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights, thereby realizing the adaptive dynamic allocation of multi-objective optimization weights. S114: Based on the customized bus route planning basic dataset, the weighted dynamic hypergraph structure, and the dynamic weight coefficients, construct a multi-objective optimization overall objective function and inject hard constraints to form a complete multi-objective mixed integer programming model; S115: Use an integer programming solver to solve the multi-objective mixed integer programming model to obtain multiple initial route schemes. Calculate the core operation indicators of each initial route scheme and complete compliance verification. After eliminating invalid schemes, output a pool of customized bus candidate route schemes.

2. The intelligent route planning method for customized buses based on route feedback collection according to claim 1, characterized in that, The steps of acquiring customized public transport travel demand data and matching multimodal environment data to construct an original spatiotemporal demand database, and performing outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database based on the multimodal environment data to obtain a cleaned spatiotemporal demand dataset include: The customized bus travel demand data is obtained through the user terminal. Each piece of customized bus travel demand data includes the unique demand identifier, the latitude and longitude coordinates of the boarding point, the latitude and longitude coordinates of the alighting point, the user's expected boarding time window, and the user's expected alighting time window. Acquire the multimodal environment data that matches the customized public transportation travel demand data. The multimodal environment data includes urban road network topology data, weather warning data, historical traffic status data, and special event identifier data for holidays and school opening seasons. All the acquired customized public transport travel demand data are correlated and matched with the multimodal environment data in terms of time and space. The user's expected boarding time window and the user's expected alighting time window are used as the time reference, and the boarding point latitude and longitude coordinates and the alighting point latitude and longitude coordinates are used as the spatial reference. The corresponding spatiotemporal range of the multimodal environment data is bound to each customized public transport travel demand data to construct the original spatiotemporal demand database. Perform anomaly removal operation: Based on the urban road network topology data, calculate the vertical distance from the latitude and longitude coordinates of each boarding point and the latitude and longitude coordinates of each alighting point to the nearest road network, and remove data with a vertical distance exceeding 500 meters. Perform duplicate request data deduplication: For multiple identical request data submitted by the same user, only one valid data is retained, and all other duplicate data is removed; Perform demand time window adjustment operation: Based on the weather warning data, if severe weather exists in the corresponding time and space range, the start time of the user's expected boarding time window will be extended forward by 10% and the end time of the user's expected alighting time window will be extended backward by 10%. Based on the historical traffic data, if the corresponding spatiotemporal range belongs to a normally congested road section, the start time of the user's expected boarding time window will be extended forward by 15% and the end time of the user's expected alighting time window will be extended backward by 15%. Based on the special event identifier data of holidays and school opening season, if the corresponding spatiotemporal range is in the peak travel period of holidays or the school opening season, the start time of the user's expected boarding time window will be extended forward by 20% and the end time of the user's expected alighting time window will be extended backward by 20%. If a single customized bus travel demand data point simultaneously meets multiple of the above correction conditions, the maximum extension duration among the conditions will be taken, without repeated addition of durations; the upper limit threshold for the extension duration is 30 minutes.

3. The customized bus intelligent route planning method based on route opinion collection according to claim 2, characterized in that: The structure of the standardized spatiotemporal object of the boarding point is as follows: ； in, This is the unique requirement identifier. Here are the longitude coordinates of the boarding point. Here are the latitude coordinates of the boarding point. This is the midpoint timestamp of the user's expected boarding time window after the demand time window has been corrected. The structure of the standardized spatiotemporal object of the drop-off point is as follows: ； in, The unique demand identifier is shared with the standardized spatiotemporal object of the boarding point. Here are the longitude coordinates of the drop-off point. Here are the latitude coordinates of the drop-off point. The timestamp of the midpoint of the user's expected drop-off time window after correction by the demand time window; The core clustering parameters include the spatial decay radius. Time decay coefficient and minimum number of cluster points The steps for adaptive matching settings of the clustering core parameters include: Based on the functional areas of the standardized spatiotemporal objects at the boarding points, they are classified into commercial and industrial parks, residential areas, and rural areas, and corresponding spatial attenuation radii are set for each. ; Based on the standardized spatiotemporal objects of the boarding points, the travel time periods are classified into weekday morning and evening peak hours, weekday off-peak hours, and holidays, and corresponding time decay coefficients are set for each. ; Based on the data scale of the standardized spatiotemporal object of the boarding point, the minimum number of cluster points is calculated. : ； in, The total number of standardized spatiotemporal objects for the boarding point within the standardized spatiotemporal object set for the boarding point. This represents the total number of pre-set customized bus routes. Based on the standardized spatiotemporal object set of boarding points and the clustering core parameters, the steps for calculating the comprehensive similarity between any two standardized spatiotemporal objects of boarding points and constructing the comprehensive similarity matrix of boarding points include: Iterate through all pairwise combinations of the standardized spatiotemporal objects of the boarding points within the set of standardized spatiotemporal objects of the boarding points. and Calculate the spatial correlation weight between the two points respectively. Time-related weights ; Among them, the spatial correlation weight : ； in, Standardize the spatiotemporal object of the boarding point and The Euclidean distance is calculated using latitude and longitude coordinates; The spatial attenuation radius; The time-related weight : ； in, The absolute difference between the timestamps of the two points; The time decay coefficient is mentioned above; Based on the spatial correlation weight Time-related weights Calculate the comprehensive similarity : ； The overall similarity The range of values ​​is A higher value indicates a stronger spatial proximity and temporal consistency between the two points, and a higher cluster correlation. Complete the comprehensive similarity of all pairwise combinations of spatiotemporal objects within the standardized spatiotemporal object set of the boarding point. After calculation, a comprehensive similarity matrix for the boarding points is constructed. This matrix is ​​a two-dimensional square matrix with equal rows and columns. Each row and column corresponds to a standardized spatiotemporal object of the boarding point, and each cell element represents the comprehensive similarity score. .

4. The intelligent route planning method for customized buses based on route feedback collection according to claim 3, characterized in that, Based on the calibrated boarding point comprehensive similarity matrix and the minimum number of cluster points in the clustering core parameters, the spatiotemporal core objects and potential noise points of the boarding points are traversed, judged, and filtered. The steps for iteratively generating a spatiotemporal demand cluster for boarding points based on the core spatiotemporal object of the boarding point, and obtaining a set of effective spatiotemporal demand clusters after completing the secondary verification and filtering of the potential noise points include: S1051: Initialize the classification state of all the standardized spatiotemporal objects at the boarding points to unclassified and the processing state to unprocessed; initialize the initial value of the auto-incrementing cluster number sequence to 0; and initialize the set of potential noise points to empty. The classification status includes: unclassified, classified, and marked; where unclassified means that the object has not completed the verification and attribution determination; classified means that the object has been assigned to the determined spatiotemporal demand cluster of the boarding point; marked means that the object has completed the first round of verification and has been identified as a potential noise point, and is waiting for the second verification and filtering. S1052: Traverse each of the standardized spatiotemporal objects of the boarding point in the set of standardized spatiotemporal objects of the boarding point and check its classification status: if it is classified, continue to traverse the next standardized spatiotemporal object of the boarding point; if it is unclassified, treat it as an unclassified spatiotemporal object Pi and execute S1053. S1053: For the unclassified spatiotemporal object Pi, select the comprehensive similarity scores with the unclassified spatiotemporal object Pi from the comprehensive similarity matrix of the boarding point. All standardized spatiotemporal objects of the boarding points that are greater than or equal to the neighborhood similarity threshold constitute the spatiotemporal neighborhood point set of the unclassified spatiotemporal object Pi; the neighborhood similarity threshold is a preset configurable parameter; S1054: Count the total number of standardized spatiotemporal objects of the boarding point within the spatiotemporal neighborhood point set. If the total number is less than the minimum clustering point number... If the unclassified spatiotemporal object Pi is determined to be a potential noise point, it is added to the potential noise point set, and its classification status is updated to marked. Then, return to S1052 to continue the traversal operation. If the total number is greater than or equal to the minimum number of cluster points If so, the unclassified spatiotemporal object Pi is determined to be the core spatiotemporal object of the boarding point, the auto-incrementing cluster number is assigned to it, and the process proceeds to S1055. S1055: All the core spatiotemporal objects of the boarding point, as well as all the standardized spatiotemporal objects of the boarding point in its spatiotemporal neighborhood point set, are classified into the spatiotemporal demand cluster of the boarding point corresponding to the auto-incrementing cluster number, and their classification status is uniformly updated to classified. S1056: Iterate through all standardized spatiotemporal objects of the current boarding point spatiotemporal demand cluster whose processing status is unprocessed. For each currently traversed cluster object Px, filter out the comprehensive similarity between the cluster object Px and the current cluster object Px from the comprehensive similarity matrix of the boarding points. All standardized spatiotemporal objects of the boarding points that are greater than or equal to the neighborhood similarity threshold constitute the spatiotemporal neighborhood point set of the intra-cluster object Px; Count the total number of objects in the spatiotemporal neighborhood point set of the cluster object Px. If the total number is greater than or equal to the minimum clustering point number... If the object Px in the cluster is determined to be the new spatiotemporal core object of the boarding point, all objects in the spatiotemporal neighborhood point set whose classification status is unclassified are simultaneously assigned to the spatiotemporal demand cluster of the boarding point corresponding to the current auto-incrementing cluster number, and their classification status is updated to classified. After completing the above judgment and attribution operations, the processing status of the current cluster object Px is updated to processed; the above traversal operation is continued until the processing status of all objects in the boarding point spatiotemporal demand cluster is processed, thereby completing the generation of the boarding point spatiotemporal demand cluster, and the value of the auto-incrementing cluster number is incremented by one, and then proceeding to S1057. S1057: If there are still unclassified standardized spatiotemporal objects in the set of standardized spatiotemporal objects of the boarding point, return to S1052 to continue the traversal operation; if all the standardized spatiotemporal objects of the boarding point are classified or marked, proceed to S1058. S1058: For all potential noise points in the set of potential noise points, perform secondary verification and filtering: calculate the comprehensive similarity between the potential noise point and all generated spatiotemporal core objects of the boarding point. Take the maximum value among them; If the maximum value is greater than or equal to the noise point attribution threshold, then the potential noise point is assigned to the spatiotemporal demand cluster of the boarding point spatiotemporal core object corresponding to the maximum value, and its classification status is updated to classified; the noise point attribution threshold is a preset configurable parameter and is less than the neighborhood similarity threshold. If the maximum value is less than the noise point attribution threshold, the potential noise point is marked as a spatiotemporally isolated noise point and is not included in any of the boarding point spatiotemporal demand clusters. S1059: After completing the generation of all the aforementioned spatiotemporal demand clusters for boarding points and the filtering of noise points, the set of effective spatiotemporal demand clusters for boarding points is obtained.

5. The intelligent route planning method for customized buses based on route feedback collection according to claim 4, characterized in that, Based on the customized bus route planning dataset, the following steps are taken to define demand nodes, road nodes, environmental event nodes, and hyperedge mapping rules, construct an initial hypergraph topology, realize cross-dimensional unified association between demand, road network, and environment, and generate a weighted dynamic hypergraph structure by constructing a dynamic weight function for each hyperedge and assigning weights. These steps enable the quantitative representation of the priority and traffic constraints of different demand clusters: Define the demand node, the road node, and the environmental event node; wherein, the demand node corresponds to the spatiotemporal demand cluster, the road node corresponds to the intersection and road segment endpoint in the urban road network topology data, and the environmental event node corresponds to the road construction and road closure event data and the event node affecting traffic in the severe weather data; Each demand node is mapped to an independent hyperedge, and each hyperedge connects all the road nodes covered by the virtual site corresponding to the demand node, as well as all the environmental event nodes within the time distribution range of the demand node, thereby realizing a cross-dimensional unified association between demand, road network and environment. Based on the definitions of the demand nodes, road nodes, and environmental event nodes, as well as the hyperedge mapping rules, the initial hypergraph topology is constructed, clarifying the association relationships of all nodes and the coverage of hyperedges; Construct the dynamic weight function, which is as follows: ； in, For the current planning period; The base travel time corresponding to the demand node; This is the time sensitivity coefficient; The cluster demand of the demand node corresponding to the superedge; The average traffic congestion index of the super-edge covered road nodes is obtained through the historical traffic flow data; This is the demand weighting coefficient. It is the traffic weighting coefficient, and and The sum of is 1; Calculate the dynamic weight value of each hyperedge; the higher the dynamic weight value, the higher the priority of the corresponding spatiotemporal demand cluster, the better the passage conditions, and the more likely it is to be covered in the route planning. The dynamic weight values ​​are assigned to the corresponding hyperedges to construct the weighted dynamic hypergraph structure.

6. The intelligent route planning method for customized buses based on route feedback collection according to claim 5, characterized in that, Based on the weighted dynamic hypergraph structure, a two-layer graph convolutional network is constructed, and inter-layer convolution operations are performed by inputting the hypergraph adjacency matrix and node attribute matrix to extract the spatial association, demand intensity, and environmental constraint features of nodes, generating a set of hypergraph node embedding vectors. Based on the set of hypergraph node embedding vectors, a meta-learning network containing two fully connected layers and a normalized exponential function layer is constructed to complete the nonlinear mapping and normalization of node features to multi-objective weights, outputting dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights. The steps to achieve adaptive dynamic allocation of multi-objective optimization weights include: Extract the numbering order of all nodes and the attribute dimensions of the hypergraph nodes in the weighted dynamic hypergraph structure, and construct a two-layer graph convolutional network; set the input layer dimension of the graph convolutional network to be completely matched with the attribute dimensions of the hypergraph nodes, set the first hidden layer dimension of the graph convolutional network to 64 dimensions, the second output layer dimension to 32 dimensions, and set the activation function of both layers of the graph convolutional network to be a linear rectified function, thereby completing the initial configuration of the two-layer graph convolutional network; Based on the initial hypergraph topology and the weighted dynamic hypergraph structure, the association relationships of all nodes and the dynamic weights of the corresponding hyperedges are statistically analyzed to generate a hypergraph adjacency matrix. Based on the customized bus route planning dataset, all nodes' full-dimensional standardized attributes are extracted and arranged in the same order as the hypergraph adjacency matrix to generate a node attribute matrix. The hypergraph adjacency matrix and the node attribute matrix are input into the initialized two-layer graph convolutional network. Inter-layer convolution is performed in the first graph convolutional layer to fuse the association information between the node's own attributes and its first-order neighboring nodes, generating a 64-dimensional intermediate feature vector. This intermediate feature vector is then input into the second graph convolutional layer to perform inter-layer convolution, fusing the global association information between second-order neighboring nodes, thereby generating a corresponding 32-dimensional node embedding vector for each node. ; All node embedding vectors are arranged in the same numbering order as the hypergraph adjacency matrix and the node attribute matrix. The nodes are arranged in an ordered manner to obtain the set of embedding vectors for the hypergraph nodes; The meta-learning network is constructed, and the core weight calculation formula of the meta-learning network is as follows: ； in, It is a multilayer perceptron used to implement nonlinear mapping between node features and weights; This is a normalized exponential function used to probabilistically assign initial weights; The set of embedding vectors for the hypergraph nodes is input into the meta-learning network. First, the node features are nonlinearly mapped through two fully connected layers, transforming the high-dimensional node features into preliminary multi-objective weights. Then, the normalized exponential function layer performs normalization processing, outputting the dynamic weight coefficients optimized for multi-objective goals. These dynamic weight coefficients include path cost weights. Vehicle fixed cost weight And passenger satisfaction penalty weighting And the sum of the three is 1.

7. The intelligent route planning method for customized buses based on route feedback collection according to claim 6, characterized in that: The overall objective function for the multi-objective optimization is: ； in, The set of feasible paths For demand clusters To demand cluster The aforementioned road segment toll cost data, For decision variables, representing whether to choose arrive The path; For vehicle types, For the first The number of vehicles of this type deployed, The first of the bus vehicle operation parameters Fixed cost per vehicle for this type of vehicle; For a set of demand clusters, For the first The actual number of passengers served by each demand cluster The rated passenger capacity of a single vehicle; , , The dynamic weighting coefficient; The hard constraints include: demand coverage constraints, vehicle capacity constraints, total transport capacity constraints, route length constraints, and time window constraints.

8. A customized intelligent bus route planning system based on route feedback collection, characterized in that, include: Acquisition module: used to acquire customized public transport travel demand data and matching multimodal environmental data to build an original spatiotemporal demand database; Data cleaning module: used to perform outlier removal, duplicate demand data deduplication, and demand time window correction on the original spatiotemporal demand database to obtain a cleaned spatiotemporal demand dataset; Data standardization module: used to standardize the cleaned spatiotemporal demand dataset, and to construct a standardized spatiotemporal object for boarding point and a standardized spatiotemporal object for disembarkation point with a common unique demand identifier for each valid demand data, forming a set of standardized spatiotemporal objects for boarding point and a set of standardized spatiotemporal objects for disembarkation point respectively; Clustering parameter adaptation and similarity calculation module: used to adaptively match and set core clustering parameters based on the regional function, travel time and data scale of the standardized spatiotemporal object of the boarding point; Furthermore, based on the standardized spatiotemporal object set of the boarding point and the clustering core parameters, the comprehensive similarity of any two standardized spatiotemporal objects of the boarding point is calculated, and a comprehensive similarity matrix of the boarding point is constructed. Spatiotemporal clustering module: used to calculate the set of spatiotemporal demand clusters for valid boarding points and the set of spatiotemporal demand clusters for valid alighting points; Attribute encapsulation module: used to complete full-dimensional attribute encapsulation for all the valid boarding point spatiotemporal demand clusters and the valid alighting point spatiotemporal demand clusters respectively, to obtain fully encapsulated boarding point spatiotemporal demand clusters and fully encapsulated alighting point spatiotemporal demand clusters, thereby forming a fully encapsulated boarding point spatiotemporal demand cluster set and a fully encapsulated alighting point spatiotemporal demand cluster set. Matrix construction module: used to construct the boarding point cluster-alighting point cluster OD demand matrix, and output the set of spatiotemporal demand clusters of the fully enclosed boarding point clusters, the set of spatiotemporal demand clusters of the fully enclosed alighting point clusters, and the OD demand matrix of the boarding point cluster-alighting point clusters that can be used for route planning; The data processing module is used to construct a basic dataset for customized bus route planning based on the set of spatiotemporal demand clusters of fully encapsulated boarding points, the set of spatiotemporal demand clusters of fully encapsulated alighting points, the OD demand matrix of the boarding point cluster and alighting point cluster, and the basic data supporting route planning. The Hypergraph Topology module is used to define demand nodes, road nodes, environmental event nodes, and hyperedge mapping rules, construct the initial hypergraph topology structure, and realize cross-dimensional unified association of demand, road network, and environment. The hypergraph weight calculation module is used to construct a dynamic weight function for each hyperedge and complete the weight assignment, generating a weighted dynamic hypergraph structure to realize the quantitative representation of the priority and passage constraints of different requirement clusters. The hypergraph feature extraction module is used to construct a two-layer graph convolutional network and input the hypergraph adjacency matrix and node attribute matrix to perform inter-layer convolution operations to extract the spatial association, demand intensity and environmental constraint features of nodes and generate a set of hypergraph node embedding vectors. The weight allocation module is used to construct a meta-learning network containing two fully connected layers and a normalized exponential function layer. It completes the nonlinear mapping and normalization of node features to multi-objective weights and outputs dynamic weight coefficients including path cost weights, vehicle fixed cost weights, and passenger satisfaction penalty weights, thereby realizing the adaptive dynamic allocation of multi-objective optimization weights. The multi-objective mixed-integer programming module is used to construct the overall objective function for multi-objective optimization and inject hard constraints to form a complete multi-objective mixed-integer programming model. The scheme generation module is used to solve the multi-objective mixed integer programming model using an integer programming solver to obtain multiple initial route schemes, calculate the core operation indicators of each initial route scheme and complete compliance verification, and output a customized bus candidate route scheme pool after eliminating invalid schemes.

9. A computer device, characterized in that, include: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform the steps of the method according to any one of claims 1 to 7 when executing the instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images