A Multi-Source Data Fusion Predictive Analysis System for Urban Transportation Resilience

Abstract

This invention belongs to the field of traffic data processing technology and relates to a multi-source data fusion predictive analysis system for urban traffic resilience. The invention generates candidate connection edges based on multi-source traffic operation values ​​and a fixed spatial grid through a generation module; generates connection traffic constraint values ​​based on floor code values, elevation values, and the open status of cross-level connection facilities through a verification module; and generates sub-mode traffic values ​​based on these connection traffic constraint values ​​to identify false connection edges. A correction module generates underestimation of transfer time and overstated value for alternative routes to correct connection connectivity and alternative redundancy values. A prediction module generates traffic resilience prediction values. This invention solves the problem that fixed spatial grid fusion methods easily lead to incorrect connections between traffic connection points that are similar in planar plane but different in spatial hierarchy and lack available cross-level connection facilities, resulting in overestimation of connection connectivity and alternative route redundancy. This improves the accuracy of urban traffic resilience prediction results in multi-layered three-dimensional traffic spaces.

Description

Technical Field

[0001] This invention relates to the field of traffic data processing technology, and in particular to a multi-source data fusion predictive analysis system for urban traffic resilience. Background Technology

[0002] In intelligent transportation data processing scenarios such as urban traffic resilience assessment, emergency evacuation of transportation hubs, and optimization of public transportation connections, existing technologies typically employ fixed spatial grids and fixed time slices for data alignment to achieve unified integration of multi-source data, including motor vehicles, public transportation, slow traffic, and disturbed environments. These technologies generally divide urban traffic areas into spatial grids of preset sizes and aggregate data such as bus arrivals, subway passenger flow, shared bicycle parking, park-and-ride, road operation, and weather disturbances into the same time slice. Then, based on the traffic resources within the same spatial grid, they calculate alternative path redundancy, connection connectivity, and urban traffic resilience prediction results.

[0003] However, in real-world urban transportation scenarios, rail stations, underground shopping streets, underground parking lots, bus hubs, overpasses, and elevated drop-off areas often constitute multi-level, three-dimensional transportation spaces. Different data platforms use inconsistent floor codes for the same spatial level; for example, B1 in a subway facility ledger, LG in a commercial building system, and P1 in a parking system may not correspond to the same actual spatial level. If the system merges data solely based on planar coordinates, fixed spatial grids, or original floor codes, it is prone to incorrectly identifying transportation connection points that are planarly similar but actually located at different spatial levels and lack available cross-level connecting facilities as valid connection points. This problem leads the system to generate non-existent cross-level connection paths, including impassable underground passages, parking lot entrances, elevated drop-off areas, or bus stops as valid alternative paths, thus artificially overestimating the redundancy and connectivity of multi-modal alternative paths. In disruptive scenarios such as rainstorms and flooding, station closures, subway passenger flow restrictions, underground passage closures, fire alarm linkages, or emergency evacuations, such cross-level node misconnection problems can be rapidly amplified, causing traffic management platforms to misjudge the regional traffic resilience, resulting in insufficient connecting transport capacity, incorrect evacuation route recommendations, increased passenger congestion, and secondary congestion.

[0004] Therefore, there is an urgent need to propose a multi-source data fusion prediction and analysis system for urban traffic resilience. After initial collection using a fixed spatial grid and fixed time slices, the system can uniformly map the floor codes of different platforms and combine absolute elevation, cross-floor connection facilities, entrance and exit correspondence, and actual traffic records to verify whether there are real cross-floor connection relationships between candidate traffic connection points. This will avoid cross-floor node misconnection caused by multi-platform floor code conflicts and improve the accuracy of urban traffic resilience prediction results. Summary of the Invention

[0005] The purpose of this invention is to address the problem that in the existing multi-source traffic data fusion process, relying solely on planar coordinates, fixed spatial grids, or original floor codes to determine connection relationships can easily lead to the misidentification of traffic connection points that are close in planar location but actually located at different spatial levels and lack available cross-level connection facilities as valid connection points. This results in the overestimation of connection connectivity values, substitution redundancy values, and traffic resilience prediction values. Therefore, this invention proposes a multi-source data fusion prediction and analysis system for urban traffic resilience.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a multi-source data fusion prediction and analysis system for urban traffic resilience, comprising: a generation module, used to obtain multi-source traffic operation values ​​in each fusion cycle, generate a fixed spatial grid based on an urban geographic information map, and generate candidate connection edges based on the multi-source traffic operation values ​​and the fixed spatial grid.

[0007] The verification module is used to generate spatial level values ​​based on the floor code values ​​and elevation values ​​of the traffic connection points at both ends of the candidate connection edge, generate connection access constraint values ​​based on the spatial level values ​​and the open status of cross-floor connection facilities, generate sub-mode access values ​​based on the connection access constraint values, and identify false connection edges in the candidate connection edges based on the sub-mode access values.

[0008] The correction module is used to generate an underestimation of transfer time and a false increment for alternative routes based on the false connection edge, and to correct the connection connectivity value based on the underestimation of transfer time and the replacement redundancy value based on the false increment for alternative routes.

[0009] The prediction module is used to generate a comprehensive transportation efficiency value based on multi-source traffic operation values, and to generate a traffic resilience prediction value based on connection and connectivity correction values, alternative redundancy correction values, and comprehensive transportation efficiency values.

[0010] In summary, by employing the aforementioned multi-source data fusion predictive analysis system for urban traffic resilience, the beneficial effects of this invention are as follows: This invention uses a fixed spatial grid as the initial screening criterion for candidate connection edges through a generation module, and further determines whether the candidate connection edges have a true spatial correspondence by combining floor code values, elevation values, and spatial level values ​​through a verification module. This solves the problem in existing technologies where generating effective connection relationships solely based on planar location or a fixed spatial grid easily leads to incorrect connections between traffic connection points at different spatial levels, thereby reducing the risk of misconnection of cross-level nodes in multi-level three-dimensional transportation spaces.

[0011] This invention generates connection and passage constraint values ​​based on spatial hierarchy values ​​and the open status of cross-layer connection facilities through a verification module. It further combines traffic connection point type values ​​and facility passage type values ​​to generate sub-mode passage values. This solves the problem in existing technologies where different connection modes such as pedestrian passage, motor vehicle passage, shared bicycle connection passage, and park-and-ride vehicle passage are not distinguished, leading to some actually impassable paths being counted as passable paths. It enables candidate connection edges to be verified separately according to different sub-mode passage items, improving the accuracy of connection relationship judgment.

[0012] This invention uses a correction module to generate an underestimation of transfer time and an overestimation of the number of alternative routes based on false connection edges, and then corrects the connection connectivity value and the alternative redundancy value respectively. This solves the problem in existing traffic resilience predictions where including false connection edges in connection paths leads to an underestimation of transfer time and an overestimation of the number of alternative routes. It ensures that the traffic resilience prediction value reflects the actual passable connection relationships, improving the accuracy of traffic resilience assessment and connection resource allocation under disturbance scenarios. Attached Figure Description

[0013] Figure 1 A schematic diagram of the system of the present invention is shown;

[0014] Figure 2 A schematic diagram of the false connection edge identification process of the present invention is shown;

[0015] Figure 3 A schematic diagram of the process for generating the underestimation of transfer time and the virtual value-added of alternative routes according to the present invention is shown. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0017] The technical solution of the present invention will be described below with reference to specific embodiments.

[0018] Example 1:

[0019] See Figure 1 A multi-source data fusion predictive analysis system for urban traffic resilience includes: a generation module, used to obtain multi-source traffic operation values ​​in each fusion cycle, generate a fixed spatial grid based on the urban geographic information map, and generate candidate connection edges based on the multi-source traffic operation values ​​and the fixed spatial grid.

[0020] Considering that a fixed spatial grid can quickly aggregate transportation hubs within the same area, but the fixed spatial grid can only represent the planar proximity between transportation hubs and cannot directly prove that there is a real passable relationship between two transportation hubs, in this embodiment, the generation module only uses the fixed spatial grid as the initial screening criterion for candidate hub edges, and does not directly use candidate hub edges as valid hub edges.

[0021] In one specific embodiment, firstly, the generation module obtains multi-source traffic operation values ​​for each fusion cycle. The multi-source traffic operation values ​​include traffic connection point location values, bus arrival status values, subway passenger flow status values, shared bicycle connection status values, park-and-ride status values, road operation status values, and disturbance status values.

[0022] Then, the generation module generates fixed spatial grids based on the city geographic information map according to a preset grid scale, and configures grid number and grid boundary for each fixed spatial grid.

[0023] Specifically, the preset grid scale is determined based on the distribution density of transportation connection points, the spatial extent of transportation hubs, and the positioning error of the urban geographic information map. When the distribution density of transportation connection points is high or the spatial extent of transportation hubs is small, the preset grid scale is reduced accordingly; when the distribution density of transportation connection points is low or the spatial extent of transportation hubs is large, the preset grid scale is increased accordingly. This allows a fixed spatial grid to be used to generate candidate connection edges, while avoiding excessive false connections between planar neighbors due to an excessively large grid scale.

[0024] Subsequently, the generation module determines the planar location value of each transportation connection point based on the location value of the transportation connection point, and determines the fixed spatial grid to which each transportation connection point belongs based on the planar location value.

[0025] Furthermore, the generation module combines two traffic connection points within the same fixed spatial grid into a traffic connection point pair, and obtains the floor code value and elevation value of the two traffic connection points in the traffic connection point pair.

[0026] Finally, when a pair of transportation connection points belongs to the same fixed spatial grid, the generation module generates candidate connection edges for that pair and sends the candidate connection edges, floor codes, and elevation values ​​to the verification module, so that the verification module can further determine whether the candidate connection edges are false connection edges.

[0027] Furthermore, when a traffic connection point is located within a preset proximity range of a fixed spatial grid boundary, the generation module also acquires traffic connection points within adjacent fixed spatial grids and forms traffic connection point pairs that meet preset planar proximity conditions. This avoids missing candidate connection edges due to traffic connection points being located on either side of adjacent fixed spatial grid boundaries.

[0028] Specifically, the preset proximity range is determined based on the grid scale of the fixed spatial grid and the positioning error of the traffic connection point location value; the preset planar proximity condition includes that the planar distance between two traffic connection points is not greater than the preset connection search distance. The preset connection search distance can be determined based on the acceptable connection distance of the corresponding sub-mode traffic item or historical connection records; among them, the preset connection search distance for pedestrian traffic can be determined based on the station transfer organization rules or the historical walking connection path length, the preset connection search distance for shared bicycle connection traffic can be determined based on the shared bicycle service radius or the historical cycling connection path length, and the preset connection search distance for motor vehicle traffic and park-and-ride vehicle traffic can be determined based on the road entrance connection relationship, the parking lot entrance and exit service range, or the historical vehicle connection path length. This ensures that the candidate connection edge search range under different sub-mode traffic items can match the actual connection method.

[0029] The verification module is used to generate spatial level values ​​based on the floor code values ​​and elevation values ​​of the traffic connection points at both ends of the candidate connection edge, generate connection access constraint values ​​based on the spatial level values ​​and the open status of cross-floor connection facilities, generate sub-mode access values ​​based on the connection access constraint values, and identify false connection edges in the candidate connection edges based on the sub-mode access values.

[0030] Considering that different data platforms may use different floor coding methods, and that the same floor code value may not necessarily correspond to the same actual spatial level, directly judging connection relationships based on floor code values ​​or fixed spatial grids could easily lead to incorrect connections between traffic connection points that are close in planar location but are actually inaccessible. Therefore, in this embodiment, the verification module first generates spatial level values ​​using floor code values ​​and elevation values, then generates connection access constraint values ​​using spatial level values ​​and the open status of cross-floor connection facilities, and finally judges whether candidate connection edges are actually accessible according to different sub-mode access items.

[0031] In one specific embodiment, the verification module first obtains the floor code value and elevation value of the traffic connection points at both ends of the candidate connection edge. The floor code value is the floor identifier of the traffic connection point in its respective data platform; the elevation value is the height value of the traffic connection point relative to the same height benchmark, and the elevation value can be obtained from urban geographic information maps, traffic facility ledgers, building surveying data, or manual calibration values.

[0032] Furthermore, to avoid directly equating floor code values ​​from different data platforms, in this embodiment, the verification module does not use the floor code value as the sole basis for spatial hierarchy determination. Instead, it first converts the elevation values ​​from subway facility ledgers, commercial building systems, parking systems, and urban geographic information maps to the same elevation benchmark, and then uses the floor code value as an auxiliary identifier for verification. Specifically, when the floor code values ​​of two transportation connection points are different but the difference in elevation values ​​is not greater than a preset elevation difference, the verification module determines this situation as a result of different floor codes but the same floor, and generates the same spatial hierarchy value; when the floor code values ​​of two transportation connection points are the same but the difference in elevation values ​​is greater than a preset elevation difference, the verification module determines this situation as a result of the same floor code but different floors, and generates different spatial hierarchy values. This avoids misconnection of candidate connection edges due to different platforms mixing codes such as B1, LG, and P1.

[0033] Then, the verification module, based on a preset elevation difference, categorizes traffic connection points whose elevation difference is not greater than the preset elevation difference into the same elevation range. Specifically, the preset elevation difference is determined based on the elevation value acquisition error and the undulation range of the ground at the same level, and the preset elevation difference can be between 0.5m and 1.5m.

[0034] Subsequently, when the floor codes of the traffic connection points at both ends of the candidate connection edge are the same and belong to the same elevation range, the verification module generates the same spatial level value for the traffic connection points at both ends of the candidate connection edge; when the floor codes of the traffic connection points at both ends of the candidate connection edge are the same but belong to different elevation ranges, the verification module generates different spatial level values ​​for the traffic connection points at both ends of the candidate connection edge; when the floor codes of the traffic connection points at both ends of the candidate connection edge are different but belong to the same elevation range, the verification module generates the same spatial level value for the traffic connection points at both ends of the candidate connection edge.

[0035] Furthermore, the verification module obtains the spatial level values ​​of the traffic connection points at both ends of the candidate connection edge. When the traffic connection points at both ends of the candidate connection edge have the same spatial level value, the candidate connection edge is determined as a same-layer candidate edge, and a first connection traffic constraint value is generated based on the same-layer candidate edge; when the traffic connection points at both ends of the candidate connection edge have different spatial level values, the candidate connection edge is determined as a cross-layer candidate edge, and cross-layer connection facilities used to connect the traffic connection points at both ends of the cross-layer candidate edge are obtained; when the cross-layer candidate edge has cross-layer connection facilities and the cross-layer connection facilities are in an open state, a second connection traffic constraint value is generated based on the cross-layer candidate edge; when the cross-layer candidate edge does not have cross-layer connection facilities or the cross-layer connection facilities are in a closed state, a third connection traffic constraint value is generated based on the cross-layer candidate edge.

[0036] Specifically, the open status of cross-level connection facilities can be determined based on facility operation records, station entrance and exit opening records, access control status, manual maintenance records, or traffic management platform status values. When the above records indicate that the cross-level connection facility is currently open for passage, the cross-level connection facility is determined to be in an open state; when the above records indicate that the cross-level connection facility is closed, under maintenance, has flow restrictions, or is not open for passage, the cross-level connection facility is determined to be in a closed state.

[0037] Specifically, the first connection traffic constraint value indicates that the traffic connection points at both ends of the candidate connection edge are at the same spatial level; the second connection traffic constraint value indicates that the traffic connection points at both ends of the candidate connection edge are at different spatial levels and there are cross-level connection facilities in an open state; the third connection traffic constraint value indicates that the traffic connection points at both ends of the candidate connection edge are at different spatial levels and there are no cross-level connection facilities in an open state.

[0038] Subsequently, the verification module obtains the connection traffic constraint values ​​of the candidate connection edges and classifies pedestrian traffic, motor vehicle traffic, shared bicycle connection traffic, and park-and-ride vehicle traffic as sub-mode traffic items. The verification module also obtains the traffic connection point type values ​​at both ends of the candidate connection edge; these values ​​characterize the sub-mode traffic items allowed at each traffic connection point. When the connection traffic constraint value is the first connection traffic constraint value, the verification module determines the sub-mode traffic items allowed at both ends of the candidate connection edge as passable values, and determines the remaining sub-mode traffic items as impassable values.

[0039] Specifically, the type values ​​for transportation hubs can be obtained from transportation facility ledgers, urban geographic information maps, station operation configuration tables, parking lot entrance and exit management records, or manual calibration tables. The facility access type values ​​can be obtained from cross-level connection facility ledgers, station passage configuration tables, entrance and exit management records, or manual calibration tables. This ensures that the sub-mode access items allowed at transportation hubs and the sub-mode access items allowed through cross-level connection facilities have clear data sources.

[0040] When the connection access constraint value is the second connection access constraint value, the verification module obtains the facility access type value of the cross-layer connection facility. The facility access type value is used to characterize the sub-mode access items that the cross-layer connection facility is allowed to pass through. The sub-mode access items that are allowed to be accessed by both traffic connection points at both ends of the candidate connection edge and that are allowed to be passed through by the cross-layer connection facility are determined as passable values, and the remaining sub-mode access items are determined as inaccessible values.

[0041] When the connection access constraint value is the third connection access constraint value, the verification module will determine all mode access items as non-accessible values.

[0042] See Figure 2Finally, the verification module generates the sub-mode access values ​​for candidate connection edges based on the passable or impassable values ​​corresponding to each sub-mode access item. When all sub-mode access items corresponding to a candidate connection edge are impassable, the candidate connection edge is identified as a false connection edge; when some sub-mode access items corresponding to a candidate connection edge are passable and some sub-mode access items are impassable, the candidate connection edge is identified as a false connection edge under the sub-mode access items corresponding to the impassable values; when all sub-mode access items corresponding to a candidate connection edge are passable, the candidate connection edge is not identified as a false connection edge.

[0043] The correction module is used to generate an underestimation of transfer time and a false increment for alternative routes based on the false connection edge, and to correct the connection connectivity value based on the underestimation of transfer time and the replacement redundancy value based on the false increment for alternative routes.

[0044] Considering that when generating candidate connection edges using a fixed spatial grid, candidate connection edges that are actually impassable may be included in the connection calculation, thereby underestimating transfer time and overestimating the number of alternative routes, this embodiment does not simply delete false connection edges. Instead, it further calculates the underestimation of transfer time and the overestimation of alternative routes caused by false connection edges, and uses these calculations to correct the connection connectivity value and the replacement redundancy value, respectively.

[0045] In one specific embodiment, firstly, the correction module uses candidate connection edges as the first connection edge set and determines the grid connection time value and the number of grid backup paths between the same traffic connection point pair. Specifically, when the first connection edge set is empty, the correction module determines that there are no candidate connection edges in the current fixed spatial grid and sets the connection connectivity value and alternative redundancy value of the corresponding fixed spatial grid to 0; when the first connection edge set is not empty, the correction module continues to determine the grid connection time value and the number of grid backup paths. Specifically, the grid connection time value is the connection time value between the same traffic connection point pair in the first connection edge set, and the grid connection time value is generated based on the planar distance value of the traffic connection point pair and the preset connection speed value; specifically, the correction module divides the planar distance value of the traffic connection point pair by the preset connection speed value under the corresponding sub-mode traffic item to obtain the grid connection time value under that sub-mode traffic item; specifically, the preset connection speed value is configured according to the sub-mode traffic item. For pedestrian traffic, the preset connection speed value can be determined based on the average walking speed; for motor vehicle traffic, the preset connection speed value can be determined based on the road operation status value; for shared bicycle connection traffic, the preset connection speed value can be determined based on the shared bicycle connection status value; for park-and-ride vehicle traffic, the preset connection speed value can be determined based on the park-and-ride status value. This allows the grid connection time value to be calculated separately for different sub-mode traffic items. Specifically, the preset connection speed value can be calculated from the actual travel time of the corresponding sub-mode traffic item within the historical fusion period, or it can be determined by the preset speed table of the traffic management platform. When the number of historical traffic records reaches the preset sample size, the historical traffic records are used first to generate the preset connection speed value; the preset sample size is determined based on the fusion period length and the historical record integrity of the corresponding sub-mode traffic item; when the historical records are insufficient, the default speed value in the preset speed table is used; the number of grid backup paths is the number of different connection paths that can be formed between the same traffic connection point pair in the first connection edge set.

[0046] Then, the correction module uses candidate connecting edges that are passable and not spurious connecting edges, represented by the sub-mode pass value, as the second connecting edge set, and determines the verification connecting time value and the number of verification backup paths between the same traffic connecting point pair. Specifically, the second connecting edge set is generated according to the sub-mode pass item. When the second connecting edge set under the corresponding sub-mode pass item is empty, the correction module determines that there is no passable connecting edge under that sub-mode pass item, does not generate a verification connecting time value under that sub-mode pass item, and sets the number of verification backup paths under that sub-mode pass item to 0. When the second connecting edge set under the corresponding sub-mode pass item is not empty, the correction module continues to determine the verification connecting time value and the number of verification backup paths. Specifically, the grid connecting time value, verification connecting time value, grid backup path number value, and verification backup path number value are all determined according to the corresponding sub-mode pass item. For candidate connecting edges where some sub-mode access items are non-accessible, the candidate connecting edge will not enter the second connecting edge set under the sub-mode access item corresponding to the non-accessible value, but will enter the second connecting edge set under the sub-mode access item corresponding to the accessible value.

[0047] Subsequently, the correction module determines the verification connection time value and the number of verification backup paths based on the second set of connection edges. The verification connection time value is the connection time value between the same traffic connection point pairs in the second set of connection edges; when the second set of connection edges contains cross-level candidate edges, the verification connection time value also includes the travel time value corresponding to the cross-level connecting facility; specifically, the travel time value corresponding to the cross-level connecting facility is determined according to the facility type of the cross-level connecting facility. When the cross-level connecting facility is a staircase, escalator, underground corridor, sky bridge, or in-station transfer passage, the correction module generates the travel time value corresponding to the cross-level connecting facility based on the ratio of the connecting path length to the average passing speed; when the cross-level connecting facility is an elevator, the correction module generates the travel time value corresponding to the cross-level connecting facility based on the sum of the elevator waiting time and the elevator running time. The number of valid backup paths is determined by the number of different connection paths that can be formed between the same pair of traffic connection points in the second connection edge set. Connection paths with the same sequence of traffic connection points are counted as the same connection path. Different connection paths are counted based on whether they repeatedly pass through the same traffic connection point. When a connection path repeatedly passes through the same traffic connection point, that connection path is not included in the grid backup path count or the valid backup path count. This avoids an inflated number of backup paths due to repeated counting of loop paths.

[0048] See Figure 3Furthermore, when the verified connection time value is greater than the grid connection time value, the correction module determines the difference between the verified connection time value and the grid connection time value as the underestimation of transfer time; when the verified connection time value is not greater than the grid connection time value, the correction module determines the underestimation of transfer time as 0; when the number of grid backup routes is greater than the number of verified backup routes, the correction module determines the difference between the number of grid backup routes and the number of verified backup routes as the illusory increment of backup routes; when the number of grid backup routes is not greater than the number of verified backup routes, the correction module determines the illusory increment of backup routes as 0.

[0049] Specifically, the correction module obtains the underestimation of transfer time, the verified transfer time value, and the connection connectivity value to be corrected corresponding to the candidate connection edge. The connection connectivity value to be corrected is a connection connectivity value generated based on the first set of connection edges. The connection connectivity value to be corrected is a value between 0 and 1, used to characterize the degree of connectivity between traffic connection points in the first set of connection edges. The correction module can generate the connection connectivity value to be corrected based on the ratio of the number of traffic connection point pairs that can form connection paths to the total number of candidate traffic connection point pairs in the first set of connection edges. When the first set of connection edges is empty or there are no traffic connection point pairs that can form connection paths, the connection connectivity value to be corrected is determined to be 0. The correction module generates a connection time deviation value based on the ratio of the underestimation of transfer time to the verified transfer time value. Specifically, when the verified transfer time value is greater than 0, the correction module determines the ratio of the underestimation of transfer time to the verified transfer time value as the connection time deviation value; when the verified transfer time value is 0 or there are no verified transfer time values, the correction module determines the connection time deviation value to be 1. When the connection time deviation value is greater than 1, the connection time deviation value is corrected to 1; based on the difference between 1 and the connection time deviation value, a connection connectivity correction coefficient is generated; based on the product of the connection connectivity value to be corrected and the connection connectivity correction coefficient, the edge correction connection connectivity value corresponding to the candidate connection edge is generated.

[0050] Further, the correction module obtains the virtual value of the backup path, the number of backup paths in the grid, and the value of the redundancy to be corrected corresponding to the candidate connection edge. The value of the redundancy to be corrected is generated based on the first set of connection edges. The value of the redundancy to be corrected is a value between 0 and 1, used to characterize the redundancy of the alternative connection paths in the first set of connection edges. The correction module can generate the value of the redundancy to be corrected based on the ratio of the number of backup paths in the grid in the first set of connection edges to a preset upper limit of the number of backup paths. When the ratio is greater than 1, the value of the redundancy to be corrected is corrected to 1; when the number of backup paths in the grid is 0, the value of the redundancy to be corrected is determined to be 0. The preset upper limit of the number of backup paths can be determined based on the scale of the transportation hub, the number of transportation connection points in the fixed spatial grid, or the statistical results of the number of backup paths in the historical fusion period. Specifically, when the number of transportation connection points in the fixed spatial grid is large or the number of available backup paths in the historical fusion period is large, the preset upper limit of the number of backup paths is increased accordingly; when the number of transportation connection points in the fixed spatial grid is small or the number of available backup paths in the historical fusion period is small, the preset upper limit of the number of backup paths is decreased accordingly. This ensures that the redundancy value to be corrected reflects a reasonable upper limit for the number of backup routes within different transportation hub spaces. Specifically, when the number of traffic connection points within a fixed spatial grid is large or the number of available backup routes within a historical fusion period is large, the preset upper limit for the number of backup routes increases accordingly; when the number of traffic connection points within a fixed spatial grid is small or the number of available backup routes within a historical fusion period is small, the preset upper limit for the number of backup routes decreases accordingly. This ensures that the redundancy value to be corrected reflects a reasonable upper limit for the number of backup routes within different transportation hub spaces. The correction module generates a backup route deviation value based on the ratio of the virtual value of backup routes to the number of backup routes in the grid; specifically, when the number of backup routes in the grid is greater than 0, the correction module determines the ratio of the virtual value of backup routes to the number of backup routes in the grid as the backup route deviation value; when the number of backup routes in the grid is 0, the correction module determines the backup route deviation value as 0. When the backup path deviation value is greater than 1, the backup path deviation value is corrected to 1; based on the difference between 1 and the backup path deviation value, an alternative redundancy correction coefficient is generated; based on the product of the alternative redundancy value to be corrected and the alternative redundancy correction coefficient, the edge correction alternative redundancy value corresponding to the candidate connection edge is generated.

[0051] The correction module configures edge weight values ​​for candidate connecting edges based on the transportation connection point type value, historical connecting passenger flow, and mode-specific access items. Specifically, the larger the historical connecting passenger flow or the more than one of the transportation connection points connected to the candidate connecting edge (railway station, bus station, park-and-ride entrance), the larger the corresponding edge weight value. The correction module normalizes the edge weight values ​​of each candidate connecting edge so that the sum of the edge weight values ​​for all candidate connecting edges under the same fixed spatial grid or the same transportation connection point is 1. The normalized edge weight values ​​are values ​​between 0 and 1. When a candidate connecting edge lacks historical connecting passenger flow, the correction module configures a default edge weight value based on the transportation connection point type value and mode-specific access items, and then normalizes the default edge weight value. Subsequently, the correction module performs a weighted summation of the edge-corrected connectivity value and edge weight value corresponding to each candidate connection edge to generate a connection connectivity correction value; and performs a weighted summation of the edge-corrected alternative redundancy value and edge weight value corresponding to each candidate connection edge to generate an alternative redundancy correction value. Thus, the connection connectivity correction value and the alternative redundancy value can reflect the degree of influence of different connection edges in actual traffic transitions.

[0052] The prediction module is used to generate a comprehensive transportation efficiency value based on multi-source traffic operation values, and to generate a traffic resilience prediction value based on connection and connectivity correction values, alternative redundancy correction values, and comprehensive transportation efficiency values.

[0053] Considering that traffic resilience predictions need to simultaneously reflect traffic operation efficiency, connectivity level, and alternative path redundancy level, directly using uncorrected connectivity and alternative redundancy values ​​would subject the predicted values ​​to false connections. Therefore, in this embodiment, the prediction module uses corrected connectivity and alternative redundancy values ​​in the traffic resilience prediction process.

[0054] In one specific embodiment, the prediction module first generates a comprehensive transportation efficiency value based on bus arrival status values, subway passenger flow status values, road operation status values, shared bicycle connection status values, park-and-ride status values, and disturbance status values. Specifically, the disturbance status value is used to characterize the impact of heavy rain, station closures, subway flow restrictions, underground passage closures, or fire alarm linkage on traffic operation efficiency. When standardizing the disturbance status value, the more severe the disturbance, the smaller the standardized disturbance status value; when there is no disturbance or the disturbance is minor, the standardized disturbance status value is larger. This ensures that the standardized disturbance status value maintains the same direction as other status values, meaning that a larger value indicates a more favorable traffic operation state for maintaining traffic resilience.

[0055] Specifically, the prediction module standardizes the bus arrival status value, subway passenger flow status value, road operation status value, shared bicycle connection status value, park-and-ride status value, and disturbance status value to values ​​between 0 and 1; where a larger value indicates that the corresponding traffic operation status is more conducive to maintaining traffic resilience. Subsequently, the prediction module averages the standardized bus arrival status value, subway passenger flow status value, road operation status value, shared bicycle connection status value, park-and-ride status value, and disturbance status value to generate a comprehensive transportation efficiency value.

[0056] Specifically, the prediction module can convert each state value into a value between 0 and 1 based on a preset level mapping table, or it can perform normalization processing based on the maximum and minimum values ​​of the corresponding state values ​​within the historical fusion period. When using the preset level mapping table, different state levels correspond to preset standardized values; when using normalization processing, the current state value is mapped to the range defined by the historical minimum and maximum values.

[0057] Then, the prediction module obtains the connection connectivity correction value, the alternative redundancy correction value, and the overall transportation efficiency value.

[0058] Subsequently, the prediction module configures connection weight values ​​for the connection connectivity correction value, redundancy weight values ​​for the replacement redundancy correction value, and efficiency weight values ​​for the overall transportation efficiency value, with the sum of the connection weight value, redundancy weight value, and efficiency weight value being 1. Specifically, the connection weight value, redundancy weight value, and efficiency weight value can be determined based on the change range of the connection connectivity correction value, replacement redundancy correction value, and overall transportation efficiency value within the historical fusion period, or they can be determined by a preset weight table; when determining the weight value based on the change range, the value with the larger change range is assigned a higher weight; when the weight value is determined using a preset weight table, the prediction module reads the connection weight value, redundancy weight value, and efficiency weight value from the preset weight table.

[0059] Finally, the prediction module generates a connection resilience score based on the product of the connection connectivity correction value and the connection weight value, a redundancy resilience score based on the product of the alternative redundancy correction value and the redundancy weight value, an efficiency resilience score based on the product of the comprehensive transportation efficiency value and the efficiency weight value, and a traffic resilience prediction value based on the sum of the connection resilience score, the redundancy resilience score and the efficiency resilience score.

[0060] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the multi-source data fusion prediction and analysis system for urban traffic resilience and its inventive concept, should be covered within the scope of protection of the present invention.

Claims

1. A multi-source data fusion predictive analysis system for urban traffic resilience, characterized in that, include: The generation module is used to obtain multi-source traffic operation values ​​in each fusion cycle, generate a fixed spatial grid based on the city geographic information map, and generate candidate connection edges based on the multi-source traffic operation values ​​and the fixed spatial grid. The verification module is used to generate spatial level values ​​based on the floor code values ​​and elevation values ​​of the traffic connection points at both ends of the candidate connection edge, generate connection access constraint values ​​based on the spatial level values ​​and the open status of cross-floor connection facilities, generate sub-mode access values ​​based on the connection access constraint values, and identify false connection edges in the candidate connection edges based on the sub-mode access values. The correction module is used to generate an underestimation of transfer time and a false value-added alternative route based on the false connection edge, and to correct the connection connectivity value based on the underestimation of transfer time and correct the replacement redundancy value based on the false value-added alternative route. The prediction module is used to generate a comprehensive transportation efficiency value based on multi-source traffic operation values, and to generate a traffic resilience prediction value based on connection and connectivity correction values, alternative redundancy correction values, and comprehensive transportation efficiency values.

2. The multi-source data fusion predictive analysis system for urban traffic resilience according to claim 1, characterized in that, The methods used by the correction module to generate underestimations of transfer times and overstated values ​​for alternative routes include: The candidate connection edges are used as the first set of connection edges to determine the grid connection time value and the number of grid backup paths. The candidate connection edges that are passable and not spurious connection edges are used as the second connection edge set to determine the connection time value and the number of backup paths to verify. When the verified connection time value is greater than the grid connection time value, the difference between the verified connection time value and the grid connection time value is determined as the underestimation of the transfer time; otherwise, the underestimation of the transfer time value is determined as 0. When the number of grid backup paths is greater than the number of verification backup paths, the difference between the number of grid backup paths and the number of verification backup paths is determined as the virtual value of backup paths; otherwise, the virtual value of backup paths is determined to be 0.

3. The multi-source data fusion predictive analysis system for urban traffic resilience according to claim 1, characterized in that, The methods for generating candidate connection edges by the generation module include: Determine the planar location value of transportation connection points and their corresponding fixed spatial grids based on urban geographic information maps; Two traffic connection points within the same fixed spatial grid are paired to obtain their floor code and elevation values, and candidate connection edges are generated.

4. The multi-source data fusion predictive analysis system for urban traffic resilience according to claim 1, characterized in that, The methods used by the verification module to generate spatial hierarchy values ​​include: Obtain the floor code values ​​and elevation values ​​relative to the same height benchmark at both ends of the candidate connection edge; determine whether the traffic connection points at both ends of the candidate connection edge belong to the same elevation range based on the preset elevation difference; When the floor code values ​​are the same and belong to the same elevation range, or when the floor code values ​​are different but belong to the same elevation range, the same spatial level value is generated. When floor code values ​​are the same but belong to different elevation ranges, different spatial level values ​​are generated; When floor codes are different and belong to different elevation ranges, different spatial level values ​​are generated.

5. A multi-source data fusion predictive analysis system for urban traffic resilience according to claim 1, characterized in that, The methods used by the verification module to generate connection and access constraint values ​​include: When both ends of a candidate connection edge have the same spatial level value, a first connection passage constraint value is generated. When the two ends have different spatial level values ​​and the open status of the cross-level connection facility indicates the existence of an open cross-level connection facility, a second connection passage constraint value is generated. When the two ends have different spatial level values ​​and there is no open cross-level connection facility, a third connection access constraint value is generated.

6. The multi-source data fusion predictive analysis system for urban traffic resilience according to claim 5, characterized in that, The methods used by the verification module to generate mode-specific pass values ​​include: Pedestrian passage, motor vehicle passage, shared bicycle shuttle passage, and park-and-ride vehicle passage are classified as different modes of passage. When the connection passage constraint value is the first connection passage constraint value, the passable or impassable value of each sub-mode passage item is determined based on the traffic connection point type value used to characterize the sub-mode passage items that are allowed to access the traffic connection point. When the connection passage constraint value is the second connection passage constraint value, the passable or impassable value of each sub-mode passage item is determined based on the traffic connection point type value at both ends of the candidate connection edge and the facility passage type value used to characterize the sub-mode passage items that are allowed to pass through the cross-level connection facility. When the connection passage constraint value is the third connection passage constraint value, all mode passage items will be determined as non-passable values.

7. A multi-source data fusion predictive analysis system for urban traffic resilience according to claim 6, characterized in that, The verification module identifies spurious connection edges using the following methods: When all the passable items of a candidate connection edge are impassable, it is identified as a false connection edge. When some sub-mode access items are not accessible, they are identified as false connection edges under the corresponding sub-mode access items; When all passable items in the partial mode are passable values, it is uncertain whether it is a false connection edge.

8. A multi-source data fusion predictive analysis system for urban traffic resilience according to claim 2, characterized in that, The methods for correcting connection connectivity values ​​by the correction module include: A connection time deviation value is generated based on the ratio of the underestimation of transfer time to the verified connection time value, and connection time deviation values ​​greater than 1 are corrected to 1. A connection connectivity correction coefficient is generated based on the difference between 1 and the connection time deviation value; Based on the connection connectivity value to be corrected and the connection connectivity correction coefficient generated from the first set of connection edges, the connection connectivity correction value is generated.

9. A multi-source data fusion predictive analysis system for urban traffic resilience according to claim 2, characterized in that, The methods for correcting redundant values ​​in the correction module include: The backup path deviation value is generated based on the ratio of the virtual value of the backup path to the number of backup paths in the grid, and the backup path deviation value greater than 1 is corrected to 1. An alternative redundancy correction coefficient is generated based on the difference between 1 and the deviation value of the backup path; Based on the alternative redundancy value to be corrected and the alternative redundancy correction coefficient generated from the first set of connecting edges, the alternative redundancy correction value is generated.

10. A multi-source data fusion predictive analysis system for urban traffic resilience according to claim 1, characterized in that, The methods used by the prediction module to generate traffic resilience predictions include: Configure connection weight values, redundancy weight values, and efficiency weight values ​​for connection connectivity correction values, alternative redundancy correction values, and overall transportation efficiency values, respectively; A connection resilience score is generated based on the product of the connection connectivity correction value and the connection weight value. A redundancy resilience score is generated by multiplying the alternative redundancy correction value and the redundancy weight value. An efficiency resilience score is generated by multiplying the overall transportation efficiency value and the efficiency weight value. The traffic resilience prediction value is obtained by summing the connection resilience score, redundancy resilience score and efficiency resilience score.

Images