A multi-dimensional data fusion-based travel cost dynamic management and control system and method

Abstract

This invention discloses a dynamic travel cost management system and method based on multi-dimensional data fusion, comprising: a multi-dimensional environmental data fusion module acquiring heterogeneous data and outputting a multi-dimensional environmental disturbance variable matrix; a travel demand spatiotemporal decoupling module stripping identity identifiers to generate a spatiotemporal demand vector, and converting agreement resource quotas and cancellation / rescheduling rules into resource slice vectors; a demand-resource bipartite graph topology calculation module constructing a demand-resource bipartite graph and calculating a topology stability index; an asynchronous pre-authorization and interface timing control module executing asynchronous pre-authorization when the topology stability index is lower than a preset safety threshold, generating a virtual resource lock in the system database to maintain the delayed identity binding state; and a topology reconstruction and resource fragmentation scheduling module disconnecting connection edges for topology reconnection when a node fails, or splitting the target resource slice vector for re-matching. This invention achieves dynamic pooling allocation of travel resources, reducing cancellation / rescheduling default costs and idle resource losses.

Description

Technical Field

[0001] This invention relates to the field of corporate travel management and data processing technology, specifically to a dynamic control system and method for travel costs based on multi-dimensional data fusion. Background Technology

[0002] The corporate travel cost control system is a collection of computer hardware and software used in modern corporate administration and financial management to standardize employee travel applications, approvals, and travel service bookings. The system mainly achieves online management of travel activities through digital processes, thereby helping companies control travel expenses and improve internal collaboration efficiency.

[0003] In existing technologies, corporate travel cost control systems typically interface with internal collaborative office platforms and external travel provider networks. When the control system receives a travel application submitted by the front end and authorized, it will directly send a physical booking instruction containing the employee's real identity information to the external travel provider system based on the travel time and spatial coordinates in the application form, thereby completing the real-name binding of travel service resources such as flights and hotels and the underlying fund settlement.

[0004] However, existing corporate travel cost control systems often employ static, linear triggering logic when initiating real-name booking instructions for underlying resources. This fails to effectively establish a quantitative assessment mechanism for the deterministic execution of travel itineraries. During the actual physical execution phase, travel itineraries are highly susceptible to changes due to external factors such as deteriorating weather and traffic congestion, as well as internal factors such as project delays and meeting cancellations. Because existing systems cannot integrate data on these heterogeneous disturbances and objectively calculate the probability of itinerary disruptions, they often send underlying physical booking instructions to external supplier system interfaces prematurely when itineraries are highly unstable. This rigid resource-binding timing control method inevitably incurs high costs for ticket refunds and rebookings when companies face itinerary changes, resulting in hidden losses of corporate travel funds. Summary of the Invention

[0005] The technical problem this invention aims to solve is that existing enterprise travel cost control systems use static, linear triggering logic when triggering real-name booking instructions for underlying resources. This fails to establish a quantitative assessment mechanism for the deterministic execution of travel itineraries, and cannot integrate data on sudden changes in the external environment and internal business changes to objectively calculate the probability of itinerary disruptions. Consequently, the system sends physical booking instructions to external supplier system interfaces prematurely when itineraries are unstable, resulting in cancellation and rescheduling default costs and hidden financial losses. This invention provides a dynamic travel cost control system and method based on multi-dimensional data fusion.

[0006] The first aspect of this invention provides a dynamic management and control system for travel costs based on multi-dimensional data fusion, including a multi-dimensional environmental data fusion module, a travel demand spatiotemporal decoupling module, a demand-resource bipartite graph topology calculation module, an asynchronous pre-authorization and interface timing control module, and a topology reconstruction and resource fragmentation scheduling module;

[0007] The multidimensional environmental data fusion module acquires project scheduling data, meeting schedule data, weather warning data, traffic hub traffic data, and real-time supplier quotation data, and performs structured alignment processing on the heterogeneous data to output a multidimensional environmental disturbance variable matrix;

[0008] The travel demand spatiotemporal decoupling module receives travel application data, strips the physical entity identity data from the travel application data to generate a spatiotemporal demand vector without identity identifiers, and at the same time obtains agreement resource quota data and refund and change rule data and converts them into resource slice vectors.

[0009] The demand-resource bipartite graph topology calculation module constructs a demand-resource bipartite graph using various spatiotemporal demand vectors and resource slice vectors as nodes, and calculates the matching edge weights between nodes. The calculation process of the matching edge weights includes obtaining the spatial distance penalty value between the spatial coordinates of the target location and the spatial coordinates of the resource, obtaining the time offset penalty value between the earliest departure time and the effective time window, and combining the basic procurement cost parameters to obtain the matching edge weight value by weighted summation of the spatial distance penalty value, the time offset penalty value, and the basic procurement cost parameters.

[0010] The demand-resource bipartite graph topology calculation module combines a multidimensional environmental disturbance variable matrix to calculate the topology stability index of each demand node. The calculation process of the topology stability index includes obtaining the probability estimate of the external environment causing the journey to be blocked, the probability estimate of the internal coordination system triggering the demand cancellation, and the historical change frequency estimate. The above three estimates are weighted and summed, and the topology stability index is obtained by subtracting the weighted sum value from the value.

[0011] When the topology stability index is lower than the preset security threshold, the asynchronous pre-authorization and interface timing control module generates a data packet with a virtual placeholder identifier and sends it to the external supplier system interface to perform asynchronous pre-authorization operation. A virtual resource lock is generated in the system database. The virtual placeholder identifier is generated by concatenating the system timestamp and the business transaction number into a hash function.

[0012] The asynchronous pre-authorization and interface timing control module calculates the latest security binding timestamp based on the physical trip start time minus the supplier protocol grace time and the system interface interaction delay time.

[0013] When a node corresponding to a specific spatiotemporal demand vector fails, the topology reconstruction and resource fragmentation scheduling module disconnects the connecting edges in the demand-resource bipartite graph and releases the corresponding virtual resource lock, then searches for remaining active nodes to reconnect the topology. When a homogeneous mapping connection cannot be formed, the topology reconstruction and resource fragmentation scheduling module splits the target resource slice vector into multiple sub-resource nodes along the time axis and re-matches the nodes. The splitting operation satisfies the condition that the sum of the time spans of the split sub-resource nodes equals the total time span of the original target resource slice vector.

[0014] A second aspect of this invention provides a method for dynamic management and control of travel costs based on multi-dimensional data fusion, applied to the aforementioned dynamic management and control system for travel costs based on multi-dimensional data fusion, comprising the following processing steps:

[0015] Step 1: The multidimensional environmental data fusion module acquires heterogeneous data, performs structured alignment processing, and outputs a multidimensional environmental disturbance variable matrix;

[0016] Step 2: The travel demand spatiotemporal decoupling module strips the physical entity identity data from the travel application data to generate an identity-free spatiotemporal demand vector, and converts the agreement resource quota data and refund / change rule data into resource slice vectors;

[0017] Step 3: The demand-resource bipartite graph topology calculation module constructs a demand-resource bipartite graph using the spatiotemporal demand vector and the resource slice vector, calculates the matching edge weights, and calculates the topology stability index of the nodes corresponding to the spatiotemporal demand vector based on the multidimensional environmental disturbance variable matrix.

[0018] Step 4: When the topology stability index is lower than the preset security threshold, the asynchronous pre-authorization and interface timing control module generates a data packet with a virtual placeholder identifier to perform asynchronous pre-authorization operation and generates a virtual resource lock in the system database.

[0019] Step 5: Before the system time reaches the latest safe binding timestamp, when the node corresponding to a specific spatiotemporal demand vector fails, the topology reconstruction and resource fragmentation scheduling module disconnects the connection edge in the demand resource bipartite graph and releases the corresponding virtual resource lock for topology reconnection. When it is impossible to form an isomorphic mapping connection, the target resource slice vector is split into multiple sub-resource nodes along the time axis and node matching is performed again.

[0020] Step Six: When the system time reaches the latest security binding timestamp, or the topology stability index reaches the preset absolute security threshold, the asynchronous pre-authorization and interface timing control module extracts the real physical entity identity data and sends a data packet containing the real physical entity identity data to the external supplier system interface to execute the final confirmation instruction.

[0021] The present invention, by adopting the above technical solution, can bring the following beneficial effects:

[0022] 1. This invention extracts heterogeneous data such as weather warnings, traffic flow, and internal scheduling into a multidimensional environmental disturbance variable matrix through a multidimensional environmental data fusion module. It then uses a demand-resource bipartite graph topology calculation module to calculate the topological stability index of the nodes corresponding to the spatiotemporal demand vectors in combination with the multidimensional environmental disturbance variable matrix. This enables an objective quantitative assessment of the probability of travel schedules being blocked due to sudden environmental changes or business changes during the physical execution phase. This provides an accurate data calculation benchmark for the subsequent execution of differentiated resource binding timing control in the underlying system.

[0023] 2. This invention performs asynchronous pre-authorization by sending a data packet with a virtual placeholder identifier through an asynchronous pre-authorization module when the topology stability index is lower than the safety threshold. It establishes a virtual resource lock in the database to maintain the delayed identity binding state, and extracts the real identity data and sends the final confirmation data packet when the clock polling condition is met. This achieves the goal of locking travel resource inventory in advance while delaying the underlying real-name authentication action, thereby avoiding the cost of cancellation and rescheduling due to premature establishment of physical contractual relationship.

[0024] 3. This invention releases virtual resource locks and performs topology reconnection by topology reconstruction and resource fragmentation scheduling modules when demand nodes fail and executes a local isomorphic optimization algorithm in the bipartite graph. When there are no available matching nodes, the target resource slice vector is divided into multiple sub-resource nodes along the time axis for rematching. This realizes the dynamic flow and pooling allocation of idle travel resources within the management and control system, avoiding the direct triggering of physical cancellation instructions from external systems due to local business conflicts. This effectively reduces the sunk costs and idle losses of corporate travel resources. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the main process of the dynamic management and control method for travel costs based on multi-dimensional data fusion of the present invention;

[0026] Figure 2 This is a schematic diagram of the topology reconstruction logic of the bipartite graph of the required resources in this invention;

[0027] Figure 3 This is a structural block diagram of the travel cost dynamic control device of the present invention;

[0028] Figure 4 This is a schematic diagram of the structure of the electronic device of the present invention. Detailed Implementation

[0029] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] See attached document Figure 1-4 This invention provides a dynamic management and control system for travel costs based on multi-dimensional data fusion, comprising: a multi-dimensional environmental data fusion module 10, a travel demand spatiotemporal decoupling module 20, a demand-resource bipartite graph topology calculation module 30, an asynchronous pre-authorization and interface timing control module 40, and a topology reconstruction and resource fragmentation scheduling module 50;

[0031] The multidimensional environmental data fusion module 10 communicates with the enterprise's internal collaborative office system and external meteorological data server, transportation hub traffic server and supplier quotation system. The multidimensional environmental data fusion module 10 acquires project scheduling data, meeting schedule data, meteorological early warning data, transportation hub traffic data and supplier real-time quotation data, and performs structured alignment processing on the above heterogeneous data to output a multidimensional environmental disturbance variable matrix.

[0032] The travel demand spatiotemporal decoupling module 20 is connected to the multidimensional environmental data fusion module 10. The travel demand spatiotemporal decoupling module 20 receives travel application data sent by the front-end business system. The travel demand spatiotemporal decoupling module 20 removes the physical entity identity data from the travel application data and converts the travel application data into a spatiotemporal demand vector without identity. The spatiotemporal demand vector includes spatial coordinate information, time window constraint information, service level information, and elasticity tolerance parameters. The travel demand spatiotemporal decoupling module 20 also obtains the enterprise's agreement resource quota data and refund and change rule data in various supplier systems. The travel demand spatiotemporal decoupling module 20 converts the agreement resource quota data and refund and change rule data into resource slice vectors.

[0033] The demand-resource bipartite graph topology calculation module 30 is connected to the travel demand spatiotemporal decoupling module 20 and the multidimensional environmental data fusion module 10, respectively. The demand-resource bipartite graph topology calculation module 30 constructs a demand-resource bipartite graph using each spatiotemporal demand vector as the left node and each resource slice vector as the right node. The formula for calculating the matching edge weights in the demand-resource bipartite graph topology calculation module 30 is as follows:

[0034] ,

[0035] in, For the nodes corresponding to the spatiotemporal demand vector Nodes corresponding to resource slice vectors Edge weights between matches; The spatial coordinates of the target location Resource-bound spatial location Distance function between spaces; Earliest departure timestamp With the effective time window of resources Time window penalty function between; Based on basic procurement cost parameters; , , The normalized weighting coefficients are set.

[0036] The demand-resource bipartite graph topology calculation module 30 receives the multidimensional environmental disturbance variables output by the multidimensional environmental data fusion module 10. The formula for calculating the topology stability index by the demand-resource bipartite graph topology calculation module 30 is as follows:

[0037] ,

[0038] in, For the nodes corresponding to the spatiotemporal demand vector The topological stability index; Estimate the probability that external factors will cause travel disruptions; Estimate the probability of request cancellation triggered by the internal collaboration system; Estimating the frequency of historical changes; , , These are the corresponding weighting coefficients;

[0039] The asynchronous pre-authorization and interface timing control module 40 is connected to the demand-resource bipartite graph topology calculation module 30. The asynchronous pre-authorization and interface timing control module 40 compares the topology stability index with the preset security threshold. When the topology stability index is lower than the preset security threshold, the asynchronous pre-authorization and interface timing control module 40 generates a data packet with a virtual placeholder identifier. The asynchronous pre-authorization and interface timing control module 40 sends the data packet with the virtual placeholder identifier to the external supplier system interface to perform asynchronous pre-authorization operation and generate a virtual resource lock in the system database.

[0040] The formula for calculating the latest secure binding timestamp in the asynchronous pre-authorization and interface timing control module 40 is as follows:

[0041] ,

[0042] in, For the latest secure binding timestamp; For the node corresponding to the resource slice vector The physical travel start time; Granting a grace period to the supplier agreement; This refers to the system interface interaction delay time.

[0043] The topology reconstruction and resource fragmentation scheduling module 50 is connected to the demand-resource bipartite graph topology calculation module 30 and the asynchronous pre-authorization and interface timing control module 40, respectively. Before the system time reaches the latest safe binding timestamp, when multi-dimensional environmental disturbance variables cause the node corresponding to a specific spatiotemporal demand vector to fail, the topology reconstruction and resource fragmentation scheduling module 50 disconnects the connection edge of the node corresponding to the specific spatiotemporal demand vector in the demand-resource bipartite graph and releases the corresponding virtual resource lock. The topology reconstruction and resource fragmentation scheduling module 50 searches for the remaining active nodes in the demand-resource bipartite graph for topology reconnection. When a complete isomorphic mapping connection cannot be formed, the topology reconstruction and resource fragmentation scheduling module 50 splits the target resource slice vector into multiple sub-resource nodes along the time axis and injects the multiple sub-resource nodes into the demand-resource bipartite graph for node matching again.

[0044] When the system time reaches the latest security binding timestamp, or when the topology stability index calculated by the demand resource bipartite graph topology calculation module 30 reaches the preset absolute security threshold, the asynchronous pre-authorization and interface timing control module 40 extracts the real physical entity identity data that has been successfully mapped, and sends the data packet containing the real physical entity identity data to the external supplier system interface to execute the final confirmation instruction.

[0045] This invention provides a method for dynamic management and control of travel costs based on multi-dimensional data fusion, which may include:

[0046] Step 1: The multidimensional environmental data fusion module 10 acquires project scheduling data, meeting schedule data, weather warning data, traffic hub traffic data, and real-time supplier quotation data. The multidimensional environmental data fusion module 10 performs structured alignment processing on the heterogeneous data and outputs a multidimensional environmental disturbance variable matrix.

[0047] The travel demand spatiotemporal decoupling module 20 receives travel application data sent by the front-end business system. This module removes the physical entity identification data from the travel application data, converting it into a spatiotemporal demand vector without identification features. The specific form of the spatiotemporal demand vector is as follows:

[0048] ,

[0049] in, For the spatiotemporal demand vector, The spatial coordinates of the origin, For the destination's spatial coordinates, The earliest departure time within the time window. The latest arrival time of the time window. For travel service levels, For time flexibility tolerance;

[0050] Step 2: The travel demand spatiotemporal decoupling module 20 obtains the enterprise's agreement resource quota data and refund and change rule data in each supplier system. The travel demand spatiotemporal decoupling module 20 converts the agreement resource quota data and refund and change rule data into resource slice vectors.

[0051] Step 3: The demand-resource bipartite graph topology calculation module 30 constructs a demand-resource bipartite graph with the spatiotemporal demand vector as the left node and the resource slice vector as the right node. The demand-resource bipartite graph topology calculation module 30 calculates the matching edge weights between the corresponding nodes of the spatiotemporal demand vector and the corresponding nodes of the resource slice vector. The formula for calculating the matching edge weights by the demand-resource bipartite graph topology calculation module 30 is as follows:

[0052] ,

[0053] in, For demand nodes With resource nodes The weight of the matching edges between matches This is the spatial distance penalty value. This is the time offset penalty value. For resource nodes Basic procurement cost parameters , , These are the preset normalized weighting coefficients;

[0054] The demand-resource bipartite graph topology calculation module 30 receives multi-dimensional environmental disturbance variables output by the multi-dimensional environmental data fusion module 10. Based on these multi-dimensional environmental disturbance variables, the demand-resource bipartite graph topology calculation module 30 calculates the topology stability index of the nodes corresponding to the spatiotemporal demand vector. The formula for calculating the topology stability index by the demand-resource bipartite graph topology calculation module 30 is as follows:

[0055] ,

[0056] in, It is the topological stability index. Estimate the probability of travel disruptions caused by external environmental factors. Estimate the probability of request cancellation triggered by the internal collaboration system. Estimating the frequency of historical changes, , , These are the corresponding weighting coefficients;

[0057] Step 4: The asynchronous pre-authorization and interface timing control module 40 compares the topology stability index with the preset security threshold. When the topology stability index is lower than the preset security threshold, the asynchronous pre-authorization and interface timing control module 40 generates a data packet with a virtual placeholder identifier. The asynchronous pre-authorization and interface timing control module 40 sends the data packet with the virtual placeholder identifier to the external supplier system interface to perform asynchronous pre-authorization operation and generates a virtual resource lock in the system database.

[0058] The asynchronous pre-authorization and interface timing control module 40 calculates the latest secure binding timestamp. The formula for calculating the latest secure binding timestamp is as follows:

[0059] ,

[0060] in, To securely bind the timestamp at the latest, This represents the physical travel start time of the node corresponding to the resource slice vector. To provide a grace period for the supplier agreement, This refers to the system interface interaction delay time.

[0061] Step 5: Before the system time reaches the latest secure binding timestamp, when the multidimensional environmental disturbance variables cause the node corresponding to the specific spatiotemporal demand vector to fail, the topology reconstruction and resource fragmentation scheduling module 50 disconnects the connection edge of the node corresponding to the specific spatiotemporal demand vector in the demand-resource bipartite graph, the topology reconstruction and resource fragmentation scheduling module 50 releases the corresponding virtual resource lock, and the topology reconstruction and resource fragmentation scheduling module 50 searches for the remaining active nodes in the demand-resource bipartite graph to reconnect the topology;

[0062] When a homogeneous mapping connection cannot be formed, the topology reconstruction and resource fragmentation scheduling module 50 splits the target resource slice vector into multiple sub-resource nodes along the time axis. The topology reconstruction and resource fragmentation scheduling module 50 then injects the multiple sub-resource nodes into the demand resource bipartite graph and performs node matching again.

[0063] Step Six: When the system time reaches the latest security binding timestamp, or when the topology stability index calculated by the demand resource bipartite graph topology calculation module 30 reaches the preset absolute security threshold, the asynchronous pre-authorization and interface timing control module 40 extracts the real physical entity identity data that has been successfully mapped, and sends the data packet containing the real physical entity identity data to the external supplier system interface to execute the final confirmation instruction.

[0064] Furthermore, the processing steps for multidimensional environmental data acquisition and feature cleaning can include:

[0065] The multi-dimensional environmental data fusion module 10 establishes a communication connection with the enterprise's internal collaborative office system through the enterprise intranet data interface. The multi-dimensional environmental data fusion module 10 extracts project scheduling data and meeting schedule data from the enterprise's internal collaborative office system. The project scheduling data includes project start timestamp, project delivery timestamp, and on-site personnel requirement parameters. The meeting schedule data includes meeting geographic coordinates, unique identification codes of participants, and meeting time intervals.

[0066] The multi-dimensional environmental data fusion module 10 establishes data connections with external meteorological data servers, transportation hub traffic servers, and supplier quotation systems through public network interfaces. The multi-dimensional environmental data fusion module 10 obtains meteorological warning data of the target travel destination from the external meteorological data server. The meteorological warning data includes the predicted rainfall value and weather disaster level parameters within a specific spatial coordinate area. The multi-dimensional environmental data fusion module 10 obtains transportation hub traffic data from the transportation hub traffic server. The transportation hub traffic data includes the delay rate of inbound and outbound transportation vehicles and the passenger congestion index of the station. The multi-dimensional environmental data fusion module 10 obtains real-time quotation data from the supplier quotation system. The real-time quotation data includes the real-time inventory balance and benchmark price change ratio of various travel service resources.

[0067] The multidimensional environmental data fusion module 10 performs structured alignment processing on the acquired heterogeneous data. The multidimensional environmental data fusion module 10 extracts the time field from the heterogeneous data and uniformly parses and converts the time fields of different formats into standard timestamps with preset time standards. The multidimensional environmental data fusion module 10 extracts the spatial location field from the heterogeneous data and uniformly maps and converts the address information with text descriptions and spatial location fields of different coordinate systems into a set of standard latitude and longitude coordinates.

[0068] The multidimensional environmental data fusion module 10 determines whether there are missing fields in the structured alignment data. When missing fields exist, the multidimensional environmental data fusion module 10 uses a linear interpolation formula to complete the numerical values. The linear interpolation formula used by the multidimensional environmental data fusion module 10 is as follows:

[0069] ,

[0070] in, To fill in missing values ​​for fields. For the timestamp corresponding to the missing field, The timestamp of the previous valid data in the missing field. The timestamp of the next valid data point after the missing field. The value of the previous valid data. The value of the next valid data;

[0071] The multidimensional environmental data fusion module 10 performs deduplication on the data based on timestamps and spatial coordinates, eliminating redundant and duplicate data packets generated during network transmission. The module then merges and encapsulates the cleaned, time-aligned, and spatially aligned data into a multidimensional environmental perturbation variable matrix. The structure of this matrix is ​​as follows:

[0072] ,

[0073] in, For a multidimensional environmental disturbance variable matrix, For project scheduling data feature vectors, For the meeting schedule data feature vector, This is the feature vector of meteorological early warning data. This is a feature vector for traffic flow data at transportation hubs. Feature vectors for real-time supplier quote data;

[0074] The multidimensional environmental data fusion module 10 transmits the multidimensional environmental disturbance variable matrix to the demand-resource bipartite graph topology calculation module 30. The multidimensional environmental disturbance variable matrix serves as the input parameter for the subsequent calculation of the topology stability index by the demand-resource bipartite graph topology calculation module 30.

[0075] Furthermore, the process of de-identifying and vectorizing travel needs can include:

[0076] The travel demand spatiotemporal decoupling module 20 receives travel application data sent by the front-end business system through the enterprise's internal network. The travel application data includes the applicant's identity information, business association information, and physical itinerary information. The applicant's identity information includes the employee's ID number, ID card number, and passport number. The business association information includes the reason for the business trip and the cost center code. The physical itinerary information includes the expected departure point, destination, departure time range, and arrival time range.

[0077] The travel demand spatiotemporal decoupling module 20 performs a physical entity identification data stripping operation on the travel application data. The travel demand spatiotemporal decoupling module 20 matches the employee ID, ID card number, and passport number fields in the travel application data using regular expressions. The travel demand spatiotemporal decoupling module 20 deletes the matched employee ID, ID card number, and passport number fields from the travel application data. The travel demand spatiotemporal decoupling module 20 establishes an association mapping table between physical entity identification data and business application number. The travel demand spatiotemporal decoupling module 20 stores the association mapping table in the underlying encrypted database.

[0078] The travel demand spatiotemporal decoupling module 20 extracts physical itinerary information from the travel application data after stripping identity identifiers. The travel demand spatiotemporal decoupling module 20 converts the expected departure point in the physical itinerary information into departure point spatial coordinates. The travel demand spatiotemporal decoupling module 20 converts the destination in the physical itinerary information into destination spatial coordinates. The travel demand spatiotemporal decoupling module 20 extracts the departure time range and arrival time range in the physical itinerary information. The travel demand spatiotemporal decoupling module 20 converts the departure time range and arrival time range into the earliest departure time and latest arrival time in standard timestamp format. The earliest departure time and the latest arrival time together constitute the time window constraint parameters.

[0079] The travel demand spatiotemporal decoupling module 20 calls the pre-stored permission mapping matrix. The travel demand spatiotemporal decoupling module 20 uses the cost center code in the business association information as an index to input the permission mapping matrix. It then uses the matrix to look up and output the corresponding numerical travel service level. The travel demand spatiotemporal decoupling module 20 extracts the business trip reason tag from the business association information. Based on the business trip reason tag, the travel demand spatiotemporal decoupling module 20 matches the corresponding benchmark coefficient in the preset tolerance benchmark coefficient table.

[0080] The formula for calculating the time flexibility tolerance of the travel demand spatiotemporal decoupling module 20 is as follows:

[0081] ,

[0082] in, For time flexibility tolerance, The tolerance baseline coefficient for matching the reason for business trip. The system's preset standard buffer time constant;

[0083] The travel demand spatiotemporal decoupling module 20 combines the spatial coordinates of the departure point, the spatial coordinates of the destination, the earliest departure time, the latest arrival time, the numerical travel service level, and the time flexibility tolerance into an array to generate a spatiotemporal demand vector without identity markers. The formula for generating the spatiotemporal demand vector by the travel demand spatiotemporal decoupling module 20 is as follows:

[0084] ,

[0085] in, For the spatiotemporal demand vector, The spatial coordinates of the origin, For the destination's spatial coordinates, The earliest departure time within the time window. The latest arrival time of the time window. To quantify travel service levels, For time flexibility tolerance;

[0086] The travel demand spatiotemporal decoupling module 20 sends the completed spatiotemporal demand vector to the demand-resource bipartite graph topology calculation module 30. The spatiotemporal demand vector serves as the left input node data for the demand-resource bipartite graph topology calculation module 30 to subsequently construct the demand-resource bipartite graph.

[0087] Furthermore, the process for resource pool slicing and virtualization mapping can include:

[0088] The travel demand time-space decoupling module 20 obtains the enterprise's agreement resource quota data and refund and change rule data in each supplier system through the external supplier interface. The agreement resource quota data includes the remaining amount of flight seats, the remaining amount of hotel rooms, and the balance of the enterprise's refund and change fund pool. The refund and change rule data includes the free name change time threshold set by the supplier, the refund fee rate tier parameters, and the change fee rate tier parameters.

[0089] The travel demand spatiotemporal decoupling module 20 performs physical dimension splitting processing on the agreement resource quota data. The travel demand spatiotemporal decoupling module 20 decomposes the batch of agreement resource quota data into independent resource slice entities. The travel demand spatiotemporal decoupling module 20 assigns a globally unique resource identification code to each resource slice entity.

[0090] The travel demand spatiotemporal decoupling module 20 extracts the service attribute features of the resource slice entities, converts the service attribute features into resource type codes that the system can recognize, and extracts the geographical location information bound to the resource slice entities and converts the geographical location information into resource space coordinates in a standard format.

[0091] The travel demand spatiotemporal decoupling module 20 calculates the effective time window parameters for resource slice entities based on the free name change time threshold in the refund and rescheduling rules data and the physical trip start and end times of the resource slice entities. The effective time window parameters define the time boundaries within the system that allow node mapping resets for resource slice entities. The formula for calculating the effective time window parameters by the travel demand spatiotemporal decoupling module 20 is as follows:

[0092] ,

[0093] in, For effective time window parameters, The physical journey start time for the resource slice entity. This refers to the free name change time threshold in the refund and rescheduling rules data;

[0094] The travel demand spatiotemporal decoupling module 20 extracts the benchmark procurement price from the agreement resource quota data. This module sets the benchmark procurement price as the basic procurement cost parameter. It also reads the ticket refund fee tier parameter from the refund and change rule data and, combined with the difference between the current system time and the physical trip start and end times, calculates the penalty cost parameter for the resource slice entity. The formula for calculating the penalty cost parameter by the travel demand spatiotemporal decoupling module 20 is as follows:

[0095] ,

[0096] in, For the penalty cost parameter, Based on basic procurement cost parameters, This is a step function for the ticket refund fee rate. The physical journey start time for the resource slice entity. The current system time;

[0097] The travel demand spatiotemporal decoupling module 20 combines resource identification codes, resource type codes, resource spatial coordinates, effective time window parameters, basic procurement cost parameters, and penalty cost parameters into arrays to generate a resource slice vector. The formula for generating the resource slice vector by the travel demand spatiotemporal decoupling module 20 is as follows:

[0098] ,

[0099] in, For resource slice vectors, For resource identification code, Encoding resource types, For resource space coordinates, For effective time window parameters, Based on basic procurement cost parameters, This refers to the parameter for penalty cost;

[0100] The travel demand spatiotemporal decoupling module 20 stores the generated resource slice vectors in the underlying system memory database. The resource slice vectors are collected in the database to form a spatiotemporal resource pool. The travel demand spatiotemporal decoupling module 20 transmits the resource slice vectors to the demand resource bipartite graph topology calculation module 30. The resource slice vectors serve as the right-side input node data for the demand resource bipartite graph topology calculation module 30 to subsequently construct the demand resource bipartite graph.

[0101] Furthermore, the processing steps for the dynamic weighted bipartite graph generation mechanism may include:

[0102] The demand-resource bipartite graph topology calculation module 30 receives the set of spatiotemporal demand vectors transmitted by the travel demand spatiotemporal decoupling module 20. The demand-resource bipartite graph topology calculation module 30 defines the set of spatiotemporal demand vectors as the set of left nodes. The demand-resource bipartite graph topology calculation module 30 also receives the set of resource slice vectors transmitted by the travel demand spatiotemporal decoupling module 20. The demand-resource bipartite graph topology calculation module 30 defines the set of resource slice vectors as the set of right nodes.

[0103] The Demand-Resource Bipartite Graph Topology Calculation Module 30 constructs a demand-resource bipartite graph based on the left and right node sets. The topology of the demand-resource bipartite graph is represented as follows: ,in For the set of nodes on the left, For the set of nodes on the right, This is the set of potential matching edges connecting the left and right nodes;

[0104] For any node corresponding to a spatiotemporal demand vector in the left node set and any node corresponding to a resource slice vector in the right node set, the demand-resource bipartite graph topology calculation module 30 performs hard constraint condition verification. The hard constraint conditions include service level matching rules and spatial physical reachability matching rules. The demand-resource bipartite graph topology calculation module 30 compares the travel service level in the spatiotemporal demand vector with the resource type code in the resource slice vector. When the travel service level and the resource type code do not match, the demand-resource bipartite graph topology calculation module 30 assigns the weight of the matching edge between the corresponding nodes to infinity and removes the corresponding potential matching edge in the topology structure of the demand-resource bipartite graph.

[0105] When the travel service level matches the resource type code, the demand-resource bipartite graph topology calculation module 30 extracts the spatial coordinate information and time window constraint information from the spatiotemporal demand vector. Simultaneously, the demand-resource bipartite graph topology calculation module 30 extracts the resource spatial coordinates and effective time window parameters from the resource slice vector.

[0106] The demand-resource bipartite graph topology calculation module 30 uses the Euclidean distance formula to calculate the spatial distance penalty value between the destination spatial coordinates and the resource spatial coordinates. The formula for calculating the spatial distance penalty value by the demand-resource bipartite graph topology calculation module 30 is as follows:

[0107] ,

[0108] in, This is the spatial distance penalty value. and Demand nodes The longitude and latitude of the destination's spatial coordinates. and Resource nodes The longitude and latitude of the resource spatial coordinates;

[0109] The demand-resource bipartite graph topology calculation module 30 uses the absolute value of the time difference to calculate the time offset penalty value between the earliest departure time of the time window and the effective time window parameters. The formula for calculating the time offset penalty value by the demand-resource bipartite graph topology calculation module 30 is as follows:

[0110] ,

[0111] in, This is the time offset penalty value. For demand nodes The earliest departure time within the time window, For resource nodes Effective time window parameters;

[0112] The demand-resource bipartite graph topology calculation module 30 extracts the basic procurement cost parameters from the resource slice vector. Combining the spatial distance penalty value, time offset penalty value, and basic procurement cost parameters, the demand-resource bipartite graph topology calculation module 30 calculates the matching edge weights between corresponding nodes. The formula for calculating the matching edge weights by the demand-resource bipartite graph topology calculation module 30 is as follows:

[0113] ,

[0114] in, For demand nodes With resource nodes The weight of the matching edges between matches This is the spatial distance penalty value. This is the time offset penalty value. For resource nodes Basic procurement cost parameters , , These are the preset normalized weighting coefficients;

[0115] The demand-resource bipartite graph topology calculation module 30 traverses all node combinations in the left and right node sets to complete the weight calculation of effective potential matching edges in the demand-resource bipartite graph. The demand-resource bipartite graph topology calculation module 30 stores the demand-resource bipartite graph containing the node set and matching edge weights in the system's running memory. The demand-resource bipartite graph serves as the underlying topology data for subsequent calculation of the topology stability index and execution of topology reconstruction operations.

[0116] Furthermore, the processing steps for calculating the topological stability index may include:

[0117] The demand-resource bipartite graph topology calculation module receives the multidimensional environmental disturbance variable matrix transmitted by the multidimensional environmental data fusion module. The demand-resource bipartite graph topology calculation module extracts the meteorological warning data feature vector and the traffic hub flow data feature vector from the multidimensional environmental disturbance variable matrix. Before extraction, the two types of feature vectors are normalized to 0-1 to eliminate the difference in dimensions.

[0118] The demand-resource bipartite graph topology calculation module extracts the meteorological warning feature value corresponding to demand node i from the meteorological warning data feature vector, and extracts the transportation hub flow feature value corresponding to demand node i from the transportation hub flow data feature vector. The module then calculates the weighted sum of the meteorological warning feature value and the transportation hub flow feature value to obtain a probability estimate of whether external environmental factors cause travel disruptions. The formula for calculating this probability estimate is as follows:

[0119] ,

[0120] in, The probability estimate of the travel obstruction caused by the external environment of demand node i is 0-1; This represents the meteorological warning feature value corresponding to demand node i. The larger the value, the higher the meteorological risk. This represents the traffic flow characteristic value of the transportation hub corresponding to demand node i. The larger the value, the higher the risk of traffic congestion. and The external environment weighting coefficient is preset and assigned values ​​based on the different types of travel transportation. For air travel, the value is... =0.6、 =0.4, railway travel =0.4、 =0.6, and satisfies + =1;

[0121] The demand-resource bipartite graph topology calculation module extracts the feature vectors of project scheduling data and meeting schedule data from the multidimensional environmental disturbance variable matrix. Before extraction, the two types of feature vectors are normalized to 0-1 to unify the numerical caliber.

[0122] The demand-resource bipartite graph topology calculation module extracts the project scheduling feature value corresponding to demand node i from the project scheduling data feature vector, and extracts the meeting schedule feature value corresponding to demand node i from the meeting schedule data feature vector. The module then calculates a weighted sum of the project scheduling feature value and the meeting schedule feature value to obtain the probability estimate of demand cancellation triggered by the internal collaboration system. The formula for calculating the probability estimate by the demand-resource bipartite graph topology calculation module is as follows:

[0123] ,

[0124] in, For demand nodes The probability estimate of request cancellation triggered by the internal collaboration system, with a value ranging from 0 to 1; For demand nodes The corresponding project scheduling characteristic value; the larger the value, the higher the risk of project changes. For demand nodes The corresponding meeting schedule characteristic value; the higher the value, the higher the risk of meeting cancellation. and As a preset internal collaboration weighting coefficient, project-led travel expenses , Meeting-led business travel , And satisfy ;

[0125] The demand-resource bipartite graph topology calculation module retrieves the historical travel records of the business entity corresponding to demand node i in the database. The historical travel records select valid travel data from the past 12 months, excluding abnormal and invalid records. The module then extracts the total number of historical travels and the number of travel changes that occurred during cancellation and rescheduling operations. The formula used by the demand-resource bipartite graph topology calculation module to calculate the historical change frequency estimate is as follows:

[0126] ,

[0127] in, Estimate the historical change frequency of demand node i, with a value ranging from 0 to 1; To change the number of trips; This represents the total number of historical trips. When the total number of historical trips is 0, a value is assigned. As an intermediate reference value;

[0128] The demand-resource bipartite graph topology calculation module combines the probability estimates of travel obstruction caused by external environment, the probability estimates of demand cancellation triggered by internal coordination system, and the historical change frequency estimates to calculate the topology stability index of demand node i. The formula for calculating the topology stability index by the demand-resource bipartite graph topology calculation module is as follows:

[0129] ,

[0130] in, For demand nodes The topological stability index ranges from 0 to 1, with a larger value indicating higher stability of the required nodes. , and The normalized weighting coefficients for the preset risk dimensions are used for regular business travel. , , Emergency travel , , And satisfy ;

[0131] The demand-resource bipartite graph topology calculation module traverses all demand nodes in the left node set and completes the topology stability index calculation for all demand nodes. The demand-resource bipartite graph topology calculation module sends the calculated topology stability index to the asynchronous pre-authorization and interface timing control module. Each topology stability index serves as the basis for the asynchronous pre-authorization and interface timing control module to make subsequent judgments on timing distribution and control.

[0132] Furthermore, the processing procedure for asynchronous pre-authorization and virtual placeholder logic may include:

[0133] The asynchronous pre-authorization and interface timing control module 40 obtains the topology stability index calculated and output by the bipartite graph topology calculation module 30 of the required resources. The asynchronous pre-authorization and interface timing control module 40 reads the system's preset security threshold from the underlying database. The asynchronous pre-authorization and interface timing control module 40 compares the topology stability index with the preset security threshold.

[0134] When the topology stability index is lower than the preset security threshold, the asynchronous pre-authorization and interface timing control module 40 intercepts the physical reservation instruction for the external supplier system interface. The payload of the physical reservation instruction contains real physical entity identification data.

[0135] The asynchronous pre-authorization and interface timing control module 40 extracts the system timestamp and business transaction number of the current system. The module then concatenates the system timestamp and business transaction number into a string. Finally, the module inputs the concatenated string into a hash function to generate a virtual placeholder identifier that does not contain personal privacy information. The formula for generating the virtual placeholder identifier is as follows:

[0136] ,

[0137] in, This is a virtual placeholder identifier. For hash functions, For system timestamps Number the business transaction records. The string concatenation operator;

[0138] The asynchronous pre-authorization and interface timing control module 40 obtains the target resource slice vector matched by the requirement node in the requirement-resource bipartite graph. The asynchronous pre-authorization and interface timing control module 40 extracts the resource identification code from the target resource slice vector. The asynchronous pre-authorization and interface timing control module 40 reassembles and encapsulates the virtual placeholder identifier, resource identification code, and time window constraint information in the spatiotemporal requirement vector to generate an asynchronous pre-authorization data packet. The structure formula of the asynchronous pre-authorization data packet is as follows:

[0139] ,

[0140] in, For asynchronous preauthorization packets, This is a virtual placeholder identifier. For resource identification code, The earliest departure time within the time window. The latest arrival time of the time window;

[0141] The asynchronous pre-authorization data packet does not contain a real physical entity identity data field. The asynchronous pre-authorization and interface timing control module 40 sends the asynchronous pre-authorization data packet to the external supplier system interface to request the external supplier system interface to lock the travel service resource inventory corresponding to the resource identification code.

[0142] The asynchronous pre-authorization and interface timing control module 40 receives a resource lock confirmation receipt returned by the external supplier's system interface. Based on the resource lock confirmation receipt, the asynchronous pre-authorization and interface timing control module 40 generates a corresponding virtual resource lock in the system database. The data structure of the virtual resource lock is represented as follows:

[0143] ,

[0144] in, For virtual resource locks, For the current demand node, For the target resource slice vector, This is a virtual placeholder identifier. For delayed binding status identifier;

[0145] The virtual resource lock establishes an association mapping record between the current demand node and the target resource slice vector in the system database. The generation of the virtual resource lock causes the current demand node to enter a delayed identity binding state within the system, and the virtual placeholder operation does not complete the real-name binding authentication of the real physical entity in the external supplier system.

[0146] Furthermore, the dynamic calculation process for the latest secure binding timestamp may include:

[0147] The asynchronous pre-authorization and interface timing control module 40 extracts the time attribute field corresponding to the target resource slice vector in the underlying database, and reads the physical trip start time bound to the target resource slice vector from the time attribute field;

[0148] The asynchronous pre-authorization and interface timing control module 40 calls and parses the supplier agreement rule text through the external supplier system interface. The asynchronous pre-authorization and interface timing control module 40 extracts the free renaming deadline for the target resource slice vector from the supplier agreement rule text and converts the free renaming deadline into the supplier agreement grace period in standard time units.

[0149] The asynchronous pre-authorization and interface timing control module 40 collects the average round-trip latency of network communication output by the system network monitoring node, obtains the computation time consumed by the underlying hardware to perform encryption and encapsulation of real physical entity identity data, and reads the system's preset interface concurrent redundancy time constant.

[0150] The formula for calculating the system interface interaction delay time by the asynchronous pre-authorization and interface timing control module 40 is as follows:

[0151] ,

[0152] in, For system interface interaction delay time, The average round-trip time for network communication, The computation time required for the underlying hardware to perform the encryption and encapsulation of the real physical entity's identity data. For interface concurrency redundancy time constant;

[0153] The asynchronous pre-authorization and interface timing control module 40 calculates the latest secure binding timestamp based on the physical trip start time, the supplier agreement grace time, and the system interface interaction delay time. The formula for calculating the latest secure binding timestamp by the asynchronous pre-authorization and interface timing control module 40 is as follows:

[0154] ,

[0155] in, The latest secure timestamp is bound to the target resource slice vector j. Let j be the physical travel start time of the target resource slice vector j. The supplier agreement grace period corresponding to the target resource slice vector j. This refers to the system interface interaction delay time.

[0156] The asynchronous pre-authorization and interface timing control module 40 writes the calculated latest secure binding timestamp into the countdown polling queue in the system memory, and establishes the association between the latest secure binding timestamp and the virtual resource lock in the system memory.

[0157] The asynchronous pre-authorization and interface timing control module 40 starts the clock polling mechanism, extracts the current system timestamp, and periodically compares the current system timestamp with the latest security binding timestamp in the countdown polling queue. The latest security binding timestamp defines the absolute time boundary for maintaining the delayed identity binding state and performing internal topology reconstruction operations within the system.

[0158] Furthermore, the processing steps for internal topology reconnection and local isomorphism optimization can include:

[0159] The topology reconstruction and resource fragmentation scheduling module 50 monitors the multidimensional environmental disturbance variables output by the multidimensional environmental data fusion module 10. Within the time window when the system time has not reached the latest safe binding timestamp, if the topology stability index of the node corresponding to the specific spatiotemporal demand vector is lower than the preset failure threshold, or if the multidimensional environmental data fusion module 10 receives a clear business cancellation instruction, the topology reconstruction and resource fragmentation scheduling module 50 obtains the current connection edge of the node corresponding to the specific spatiotemporal demand vector in the demand-resource bipartite graph.

[0160] The topology reconstruction and resource fragmentation scheduling module 50 deletes the current connection edge in the demand resource bipartite graph. The topology reconstruction and resource fragmentation scheduling module 50 releases the virtual resource lock corresponding to the current connection edge in the underlying database. After the virtual resource lock is released, the original target resource slice vector is restored to an idle state node.

[0161] The topology reconstruction and resource fragmentation scheduling module 50 extracts the feature parameters of the target resource slice vector that has been restored to an idle state node. The topology reconstruction and resource fragmentation scheduling module 50 traverses the nodes corresponding to the other spatiotemporal demand vectors that are in an active state in the left node set of the demand resource bipartite graph.

[0162] The topology reconstruction and resource fragmentation scheduling module 50 calls the demand-resource bipartite graph topology calculation module 30 to calculate the matching edge weights between the nodes corresponding to each active spatiotemporal demand vector and the nodes corresponding to the target resource slice vector. The topology reconstruction and resource fragmentation scheduling module 50 executes a local isomorphism optimization algorithm to select the optimal mapping node that meets the preset tolerance threshold. The objective function and constraint formulas for the local isomorphism optimization performed by the topology reconstruction and resource fragmentation scheduling module 50 are as follows:

[0163] ,

[0164] in, The optimal mapping node number obtained through optimization. This is the set of nodes corresponding to the spatiotemporal demand vectors that are in an active state. For nodes corresponding to the spatiotemporal demand vector in an active state Nodes corresponding to the target resource slice vector in the idle state The weight of the matching edges between them The maximum matching tolerance weight threshold preset for the system;

[0165] When the local isomorphic optimization algorithm outputs a valid optimal mapping node, the topology reconstruction and resource fragmentation scheduling module 50 establishes a connection edge between the optimal mapping node and the corresponding node of the target resource slice vector in the demand resource bipartite graph. The topology reconstruction and resource fragmentation scheduling module 50 generates a new virtual resource lock in the underlying database to complete the node mapping update and topology reconnection of the physical travel resources within the system.

[0166] When there are no nodes in the active left node set that satisfy the condition that the matching edge weight is less than the maximum matching tolerance weight threshold, the topology reconstruction and resource fragmentation scheduling module 50 terminates the internal topology reconnection operation and triggers a resource fragmentation degradation instruction for the target resource slice vector.

[0167] Furthermore, the process for handling resource fragmentation, degradation, and virtual reorganization may include:

[0168] The topology reconstruction and resource fragmentation scheduling module 50 receives the resource fragmentation degradation instruction. The topology reconstruction and resource fragmentation scheduling module 50 extracts the resource type code and effective time window parameters of the target resource slice vector in the idle state. Based on the resource type code, the topology reconstruction and resource fragmentation scheduling module 50 queries the resource attribute configuration table at the bottom layer of the system. The topology reconstruction and resource fragmentation scheduling module 50 determines whether the target resource slice vector has the attribute of being divisible by the time axis.

[0169] When the target resource slice vector has a time-axis-spliable attribute, the topology reconstruction and resource fragmentation scheduling module 50 divides the target resource slice vector into multiple continuous sub-resource nodes along the time axis. The topology reconstruction and resource fragmentation scheduling module 50 performs physical duration segmentation operation according to the minimum business time granularity preset by the system. The resource segmentation operation performed by the topology reconstruction and resource fragmentation scheduling module 50 satisfies the principle of time continuity conservation. The constraint formula for the segmentation performed by the topology reconstruction and resource fragmentation scheduling module 50 is as follows:

[0170] ,

[0171] in, This is the original target resource slice vector. To divide the total number of sub-resource nodes generated, The segmented result is the first Individual resource nodes, The total time span of the original target resource slice vector. For the first The time span of each sub-resource node;

[0172] The topology reconstruction and resource fragmentation scheduling module 50 proportionally splits the basic procurement cost parameters of the original target resource slice vector according to the time span proportion of each sub-resource node. The formula for calculating the basic procurement cost parameters of the sub-resource nodes by the topology reconstruction and resource fragmentation scheduling module 50 is as follows:

[0173] ,

[0174] in, For the first Basic procurement cost parameters for each sub-resource node The basic procurement cost parameters for the original target resource slice vector;

[0175] The topology reconstruction and resource fragmentation scheduling module 50 assigns the time span and basic procurement cost parameters of the sub-resource nodes obtained from the splitting to the corresponding sub-resource nodes. The topology reconstruction and resource fragmentation scheduling module 50 generates an independent resource identifier code for each sub-resource node in the underlying database. The topology reconstruction and resource fragmentation scheduling module 50 injects each generated sub-resource node into the right node set of the demand resource bipartite graph.

[0176] The topology reconstruction and resource fragmentation scheduling module 50 deletes the right-side node corresponding to the original target resource slice vector in the demand-resource bipartite graph. The topology reconstruction and resource fragmentation scheduling module 50 calls the demand-resource bipartite graph topology calculation module 30. The demand-resource bipartite graph topology calculation module 30 calculates the matching edge weights between the node corresponding to the spatiotemporal demand vector in the active state and each newly injected sub-resource node.

[0177] The topology reconstruction and resource fragmentation scheduling module 50 performs a local isomorphic optimization operation based on the recalculated matching edge weights. When a sub-resource node successfully matches the node corresponding to the active spatiotemporal demand vector, the topology reconstruction and resource fragmentation scheduling module 50 establishes a corresponding virtual resource lock in the underlying database. The topology reconstruction and resource fragmentation scheduling module 50 records the mapping relationship between the sub-resource node and the node corresponding to the successfully matched spatiotemporal demand vector in the underlying database.

[0178] The topology reconfiguration and resource fragmentation scheduling module 50 calculates the lifecycle deadline timestamp for unmatched sub-resource nodes. The formula for calculating the lifecycle deadline timestamp by the topology reconfiguration and resource fragmentation scheduling module 50 is as follows:

[0179] ,

[0180] in, For the first The lifecycle expiration timestamp of each sub-resource node. The latest safe timestamp is bound to the original target resource slice vector. The system's preset internal processing buffer time;

[0181] The topology reconstruction and resource fragmentation scheduling module 50 periodically compares the current system timestamp with the lifecycle expiration timestamp. When the current system timestamp reaches the lifecycle expiration timestamp and the sub-resource node has not yet been matched successfully, the topology reconstruction and resource fragmentation scheduling module 50 triggers the operation of sending a physical unsubscribe instruction to the external supplier system interface.

[0182] Furthermore, the processing steps for high-priority clock polling and dual-trigger condition verification may include:

[0183] The asynchronous pre-authorization and interface timing control module 40 starts a high-priority clock polling thread. The asynchronous pre-authorization and interface timing control module 40 periodically extracts the current system timestamp through the high-priority clock polling thread at a preset frequency. The asynchronous pre-authorization and interface timing control module 40 accesses the countdown polling queue and reads the virtual resource lock that is in the delayed identity binding state.

[0184] The asynchronous pre-authorization and interface timing control module 40 extracts the latest secure binding timestamp corresponding to the virtual resource lock. The asynchronous pre-authorization and interface timing control module 40 performs the first-level time trigger condition verification. The asynchronous pre-authorization and interface timing control module 40 compares the current system timestamp with the latest secure binding timestamp. The determination formula for the first-level time trigger condition verification is:

[0185] ,

[0186] in, This is the current system timestamp. The latest security binding timestamp corresponding to the target resource slice vector j is set. When the current system timestamp is greater than or equal to the latest security binding timestamp, the first time trigger condition verification passes. The asynchronous pre-authorization and interface timing control module 40 interrupts the topology reconstruction permission for the specific virtual resource lock and triggers the physical layer identity injection operation.

[0187] The asynchronous pre-authorization and interface timing control module 40 simultaneously executes the second-level stability trigger condition verification. The demand-resource bipartite graph topology calculation module 30, based on the multi-dimensional environmental disturbance variable matrix output by the multi-dimensional environmental data fusion module 10, updates the topology stability index of demand nodes in the delayed identity binding state in real time. The asynchronous pre-authorization and interface timing control module 40 obtains the real-time updated topology stability index and retrieves the preset absolute security threshold. The determination formula for the second-level stability trigger condition verification is:

[0188] ,

[0189] in, For demand nodes The real-time topological stability index. The second stability trigger condition verification is passed when the real-time topology stability index is greater than or equal to the absolute security threshold. The asynchronous pre-authorization and interface timing control module 40 terminates the delayed identity binding state of the specific virtual resource lock in advance.

[0190] When either the first-level time-triggered condition verification or the second-level stability-triggered condition verification passes, the asynchronous pre-authorization and interface timing control module 40 extracts the node mapping relationship in the specific virtual resource lock. The asynchronous pre-authorization and interface timing control module 40 extracts the business application number bound to the demand node from the node mapping relationship. The asynchronous pre-authorization and interface timing control module 40 sends the business application number to the travel demand time-space decoupling module 20.

[0191] The travel demand spatiotemporal decoupling module 20 receives the business application number and accesses the encrypted database. The travel demand spatiotemporal decoupling module 20 extracts the real physical entity identity data corresponding to the business application number from the encrypted database. The travel demand spatiotemporal decoupling module 20 transmits the real physical entity identity data to the asynchronous pre-authorization and interface timing control module 40. The asynchronous pre-authorization and interface timing control module 40 extracts the resource identification code from the specific virtual resource lock.

[0192] The asynchronous pre-authorization and interface timing control module 40 writes the real physical entity identification data into the preset final confirmation data packet payload. The structural formula for the asynchronous pre-authorization and interface timing control module 40 to assemble the final confirmation data packet is as follows:

[0193] ,

[0194] in, To finally confirm the data packet, For resource identification code, For real physical entity identification data, For business application number;

[0195] The asynchronous pre-authorization and interface timing control module 40 sends the final confirmation data packet to the external supplier system interface to request the external supplier system interface to perform real-name binding authentication on the resource corresponding to the resource identification code using the real physical entity identity data.

[0196] Furthermore, the processing procedure for communication via the physical layer Final-Commit interface may include:

[0197] The asynchronous pre-authorization and interface timing control module 40 extracts the final confirmation data packet generated by the preceding processing. The asynchronous pre-authorization and interface timing control module 40 calls the system's underlying security encryption component and uses an asymmetric encryption algorithm to encrypt the real physical entity identification data in the final confirmation data packet. The formula for the encryption operation performed by the asynchronous pre-authorization and interface timing control module 40 is as follows:

[0198] ,

[0199] in, For encrypted entity identification data, It is an asymmetric encryption function. For real physical entity identification data, Public keys assigned to interfaces with external vendor systems;

[0200] The asynchronous pre-authorization and interface timing control module 40 establishes a data communication connection with the external supplier system interface through the encrypted transmission control protocol. The asynchronous pre-authorization and interface timing control module 40 sends the final confirmation data packet containing the encrypted entity identity data to the external supplier system interface. The asynchronous pre-authorization and interface timing control module 40 requests the external supplier system interface to perform the data writing operation of the resource identification code and the encrypted entity identity data.

[0201] The asynchronous pre-authorization and interface timing control module 40 listens to the response network port of the external supplier system interface within the set timeout threshold. The asynchronous pre-authorization and interface timing control module 40 receives the physical credential receipt data packet sent by the external supplier system interface. The asynchronous pre-authorization and interface timing control module 40 performs unpacking and field extraction operations on the physical credential receipt data packet. The structure formula of the physical credential receipt data packet is as follows:

[0202] ,

[0203] in, This is a physical credential receipt data packet. For resource identification code, Ticket confirmation codes assigned to external supplier system interfaces. For the final settlement amount parameter, Confirmation time for generating data packet stamps for external supplier system interfaces;

[0204] The asynchronous pre-authorization and interface timing control module 40 extracts the resource identification code from the physical credential receipt data packet. The asynchronous pre-authorization and interface timing control module 40 compares the resource identification code in the physical credential receipt data packet with the resource identification code in the final confirmation data packet. When the resource identification codes match, the asynchronous pre-authorization and interface timing control module 40 uses the resource identification code as the primary key to query the system database for the corresponding specific virtual resource lock.

[0205] The asynchronous pre-authorization and interface timing control module 40 writes the ticket confirmation code, final settlement amount parameter and confirmation timestamp into the extended storage field of the specific virtual resource lock. The asynchronous pre-authorization and interface timing control module 40 executes the status update instruction of the underlying database and changes the status identifier of the specific virtual resource lock from the delayed binding status to the physical binding completed status.

[0206] The asynchronous pre-authorization and interface timing control module 40 locks the topology reconstruction permission and resource fragmentation scheduling permission of a specific virtual resource lock in the underlying database. The asynchronous pre-authorization and interface timing control module 40 obtains the updated specific virtual resource lock and the associated business application number. The asynchronous pre-authorization and interface timing control module 40 sends the ticket confirmation code and final settlement amount parameters in the specific virtual resource lock to the enterprise's internal collaborative office system through the enterprise intranet data interface. The enterprise's internal collaborative office system updates the front-end business status based on the received data, realizing the synchronization of data status between the enterprise's internal collaborative office system and the external supplier system.

[0207] Furthermore, the present invention provides an electronic device that may include: a processor 601, a memory 602, a communication interface 603, and a bus 604, wherein the processor 601, the memory 602, and the communication interface 603 achieve data communication and physical electrical connection with each other through the bus 604;

[0208] The processor 601 is composed of a general-purpose central processing unit, a microprocessor, an application-specific integrated circuit, or one or more integrated circuits. The processor 601 is used to execute relevant program code to realize the dynamic management method for travel costs based on multi-dimensional data fusion provided in the embodiments of the present invention.

[0209] The memory 602 is used to store computer program code. The memory 602 includes read-only memory, random access memory, flash memory or disk storage. The memory 602 is physically and logically divided into volatile high-speed running memory and non-volatile persistent storage area. The volatile high-speed running memory is used to carry the topological structure data of the demand resource bipartite graph and the multi-dimensional environmental disturbance variable matrix in real time. The non-volatile persistent storage area is used to store encrypted real physical entity identification data and historical travel records. The computer program code stored in the memory 602 contains an instruction set for performing dynamic management of travel costs.

[0210] The communication interface 603 is used to realize data transmission between electronic devices and other external systems or hardware terminals. The communication interface 603 includes a hardware network interface controller (NIC) and a radio frequency transceiver antenna array. The communication interface 603 supports wired network communication protocols or wireless network communication protocols. Electronic devices establish data communication connections with internal enterprise collaborative office systems and external supplier system interfaces through the communication interface 603.

[0211] Bus 604 includes physical hardware cables and is used to transfer information data between processor 601, memory 602 and communication interface 603.

[0212] When the electronic device is in operation, the processor 601 calls and executes the computer program code stored in the memory 602. When the processor 601 executes the computer program code, it implements each step of the dynamic management method for travel costs based on multi-dimensional data fusion described in the foregoing embodiments. Specifically, when the processor 601 executes the computer program code, it sequentially controls and executes the operations of multi-dimensional environmental data acquisition and feature cleaning, travel demand de-identification and vectorization conversion, resource pool slicing and virtualization mapping, dynamic weighted bipartite graph generation, topology stability index calculation, asynchronous pre-authorization and virtual occupancy, internal topology reconnection and local isomorphic optimization, resource fragmentation degradation and virtual reorganization, and physical layer Final-Commit interface communication.

[0213] This invention also provides a computer-readable storage medium storing a computer program. The computer-readable storage medium includes a non-volatile storage medium and is composed of a USB flash drive, a portable hard drive, a read-only memory, a random access memory, a magnetic disk, or an optical disk.

[0214] When the processor executes the computer program, it implements the dynamic management method for travel costs based on multi-dimensional data fusion provided in the aforementioned embodiments. The instruction set contained in the computer program instructs the processor to read the multi-dimensional environmental disturbance variable matrix, construct a demand-resource bipartite graph, and perform delayed identity binding and topology reconstruction operations based on the topology stability index.

Claims

1. A multi-dimensional data fusion-based dynamic travel cost management system, characterized in that, include: The multidimensional environmental data fusion module (10) is used to acquire project scheduling data, meeting schedule data, meteorological early warning data, traffic hub traffic data and supplier real-time quotation data, and to perform structured alignment processing on heterogeneous data to output a multidimensional environmental disturbance variable matrix. The travel demand spatiotemporal decoupling module (20) is connected to the multidimensional environmental data fusion module (10) and is used to receive travel application data, strip the physical entity identity data from the travel application data to generate a spatiotemporal demand vector without identity, and obtain agreement resource quota data and refund and change rules data and convert them into resource slice vectors. The demand-resource bipartite graph topology calculation module (30) is connected to the travel demand spatiotemporal decoupling module (20) and the multidimensional environmental data fusion module (10) respectively. It is used to construct a demand-resource bipartite graph with each spatiotemporal demand vector as the left node and each resource slice vector as the right node, calculate the matching edge weight between the node corresponding to the spatiotemporal demand vector and the node corresponding to the resource slice vector, and receive the multidimensional environmental disturbance variable matrix to calculate the topological stability index of the node corresponding to the spatiotemporal demand vector. The asynchronous pre-authorization and interface timing control module (40) is connected to the demand resource bipartite graph topology calculation module (30) and is used to compare the topology stability index with the preset security threshold. When the topology stability index is lower than the preset security threshold, a data packet with a virtual placeholder is generated and sent to the external supplier system interface to perform asynchronous pre-authorization operation and generate a virtual resource lock in the system database. The topology reconstruction and resource fragmentation scheduling module (50) is connected to the topology calculation module (30) of the demand resource bipartite graph and the asynchronous pre-authorization and interface timing control module (40), respectively. It is used to disconnect the connection edge of the demand resource bipartite graph and release the corresponding virtual resource lock when the node corresponding to the demand vector in a specific time and space fails, search for the remaining active nodes to reconnect the topology, and when it is impossible to form an isomorphic mapping connection, split the target resource slice vector into multiple sub-resource nodes along the time axis and inject them into the demand resource bipartite graph to rematch the nodes.

2. The travel cost dynamic management and control system based on multi-dimensional data fusion according to claim 1, characterized in that, When the multidimensional environmental data fusion module (10) performs structured alignment processing on the acquired heterogeneous data, it extracts the time field in the heterogeneous data and uniformly parses and converts it into a standard timestamp, and extracts the spatial location field in the heterogeneous data and uniformly maps and converts it into a standard latitude and longitude coordinate set. When there are missing fields in the data after structured alignment, the multidimensional environmental data fusion module (10) uses a linear interpolation formula to complete the numerical values ​​and merges and encapsulates the cleaned, time-aligned and space-aligned data into the multidimensional environmental disturbance variable matrix.

3. The travel cost dynamic management and control system based on multi-dimensional data fusion according to claim 1, characterized in that, The travel demand spatiotemporal decoupling module (20) uses regular expressions to match and delete the employee ID, ID card number and passport number fields in the travel application data to complete the stripping operation of physical entity identity data; The travel demand spatiotemporal decoupling module (20) generates the spatiotemporal demand vector by combining arrays containing the spatial coordinates of the departure point, the spatial coordinates of the destination, the earliest departure time of the time window, the latest arrival time of the time window, the travel service level, and the time flexibility tolerance.

4. The travel cost dynamic management and control system based on multi-dimensional data fusion according to claim 1, characterized in that, The demand-resource bipartite graph topology calculation module (30) calculates the matching edge weights between the nodes corresponding to the spatiotemporal demand vector and the nodes corresponding to the resource slice vector. The formula for calculating the matching edge weights is as follows: ， in, For demand nodes With resource nodes The weight of the matching edges between matches This is the spatial distance penalty value. This is the time offset penalty value. For resource nodes Basic procurement cost parameters , , These are the preset normalized weighting coefficients.

5. The travel cost dynamic management and control system based on multi-dimensional data fusion according to claim 1, characterized in that, The demand-resource bipartite graph topology calculation module (30) calculates the topology stability index of the nodes corresponding to the spatiotemporal demand vector based on the multidimensional environmental disturbance variable matrix. The formula for calculating the topology stability index is: ， in, It is the topological stability index. Estimate the probability of travel disruptions caused by external environmental factors. Estimate the probability of request cancellation triggered by the internal collaboration system. Estimating the frequency of historical changes, , , These are the corresponding weighting coefficients.

6. The travel cost dynamic management and control system based on multi-dimensional data fusion according to claim 1, characterized in that, The asynchronous pre-authorization and interface timing control module (40) extracts the system timestamp and business serial number of the current system, concatenates them into a string, and inputs the concatenated string into a hash function to generate the virtual placeholder identifier; The asynchronous pre-authorization and interface timing control module (40) reassembles and encapsulates the virtual placeholder identifier, resource identification code and time window constraint information in the spatiotemporal demand vector to generate an asynchronous pre-authorization data packet without an internal real physical entity identity identifier data field.

7. The travel cost dynamic management and control system based on multi-dimensional data fusion according to claim 6, characterized in that, The asynchronous pre-authorization and interface timing control module (40) calculates the latest security binding timestamp, which defines the absolute time boundary for maintaining the delayed identity binding state and performing internal topology reconstruction operations within the system. The latest secure binding timestamp is calculated by subtracting the timestamp from the physical journey start time of the node corresponding to the resource slice vector, the supplier protocol grace time, and the system interface interaction delay time. The system interface interaction delay time is calculated based on the average round-trip latency of network communication, the computation time of the underlying hardware to encrypt and encapsulate the real physical entity identity data, and the interface concurrent redundancy time constant.

8. The travel cost dynamic management and control system based on multi-dimensional data fusion according to claim 6, characterized in that, The asynchronous pre-authorization and interface timing control module (40) starts a high-priority clock polling mechanism to perform the first time trigger condition verification and the second stability trigger condition verification. When the current system timestamp is greater than or equal to the latest security binding timestamp, or the real-time topology stability index is greater than or equal to the preset absolute security threshold, the asynchronous pre-authorization and interface timing control module (40) terminates the delayed identity binding state of the specific virtual resource lock in advance, extracts the real physical entity identity data that has been successfully mapped, and sends the final confirmation data packet containing the real physical entity identity data to the external supplier system interface to execute the final confirmation instruction.

9. The travel cost dynamic management and control system based on multi-dimensional data fusion according to claim 1, characterized in that, The topology reconstruction and resource fragmentation scheduling module (50) executes a local isomorphic optimization algorithm to screen the optimal mapping node that meets the preset tolerance threshold, and executes a resource fragmentation degradation instruction when it is impossible to form an isomorphic mapping connection. The topology reconstruction and resource fragmentation scheduling module (50) divides the target resource slice vector into multiple continuous sub-resource nodes along the time axis according to the minimum business time granularity preset by the system, and proportionally splits the basic procurement cost parameters of the original target resource slice vector according to the time span ratio of each sub-resource node.

10. A method for dynamic management and control of travel costs based on multi-dimensional data fusion, characterized in that, The method, applied to the multi-dimensional data fusion-based dynamic travel cost management system as described in any one of claims 1 to 9, comprises: Step 1: The multidimensional environmental data fusion module (10) acquires heterogeneous data, performs structured alignment processing, and outputs a multidimensional environmental disturbance variable matrix; Step 2: The travel demand spatiotemporal decoupling module (20) strips the physical entity identity data from the travel application data to generate a spatiotemporal demand vector without identity, and converts the agreement resource quota data and refund / change rule data into resource slice vectors; Step 3: The demand-resource bipartite graph topology calculation module (30) constructs a demand-resource bipartite graph using the spatiotemporal demand vector and the resource slice vector, calculates the matching edge weights, and calculates the topological stability index of the nodes corresponding to the spatiotemporal demand vector based on the multidimensional environmental disturbance variable matrix. Step 4: When the topology stability index is lower than the preset security threshold, the asynchronous pre-authorization and interface timing control module (40) generates a data packet with a virtual placeholder identifier to perform asynchronous pre-authorization operation and generates a virtual resource lock in the system database; Step 5: Before the system time reaches the latest safe binding timestamp, when the node corresponding to a specific spatiotemporal demand vector fails, the topology reconstruction and resource fragmentation scheduling module (50) disconnects the connection edge in the demand resource bipartite graph and releases the corresponding virtual resource lock to reconnect the topology. When it is impossible to form an isomorphic mapping connection, the target resource slice vector is split into multiple sub-resource nodes along the time axis and node matching is performed again. Step 6: When the system time reaches the latest security binding timestamp, or the topology stability index reaches the preset absolute security threshold, the asynchronous pre-authorization and interface timing control module (40) extracts the real physical entity identity data and sends a data packet containing the real physical entity identity data to the external supplier system interface to execute the final confirmation instruction.

Images