A two-layer matching-based online car-hailing aggregation platform order dispatching method, system, storage medium and electronic device

Abstract

The application provides a two-layer matching-based online car-hailing aggregation platform order dispatching method, which is used for realizing efficient driver and passenger matching in a multi-sub-platform collaborative environment. The method comprises the following steps: an aggregation platform integrates passenger order information in an order dispatching period, and broadcasts each order to a sub-platform set preferred by the passenger; each sub-platform performs local matching based on its own vehicle fleet, with the aim of maximizing utility, and returns a candidate vehicle set; the aggregation platform combines candidate results of each sub-platform, constructs a global candidate vehicle set for each passenger, and selects a vehicle with the highest utility from the global candidate vehicle set to complete final assignment. Under the premise of ensuring information isolation and operation compliance, the application realizes efficient integration of cross-platform online car-hailing resources by combining sub-platform local matching and aggregation platform global coordination, and provides reliable technical support for operation management of the online car-hailing aggregation platform.

Description

Technical Field

[0001] This invention belongs to the field of ride-hailing dispatching technology, specifically relating to a ride-hailing aggregation platform dispatching method, system, storage medium, and electronic device based on two-layer matching. Background Technology

[0002] As a new type of travel service, ride-hailing aggregation platforms integrate the transportation resources of multiple ride-hailing sub-platforms and leverage the user and technological advantages of map service providers to offer passengers a one-stop ride-hailing service. In recent years, they have experienced rapid development and have become a significant force reshaping the travel market. However, this model still faces numerous challenges in actual operation: regarding access mechanisms, some platforms are not strict enough in their qualification reviews of partners, posing compliance risks; regarding the division of responsibilities, the boundaries of rights and responsibilities between aggregation platforms and sub-platforms are not clear enough, easily leading to the shirking of responsibility; regarding service transparency, discrepancies often exist between displayed prices and actual charges, affecting user experience. Although current regulatory requirements have clearly defined the information intermediary role of aggregation platforms and restricted their direct intervention in vehicle dispatching and pricing, how to build an efficient capacity matching and resource coordination mechanism within a compliant framework remains a key technical challenge for the industry. A dynamic optimal matching method for ride-hailing aggregation platforms (application number CN202411560343.3) introduces approximate dynamic programming into the driver-passenger matching problem. It constructs a dynamic programming model with the goal of maximizing the total revenue of the aggregation platform and approximates the expected profit based on real-time supply and demand data, thereby dynamically optimizing the matching strategy. A ride-hailing platform dispatching method and system (application number CN202211239287.4) targets enterprise user scenarios. It prioritizes and reorders sub-platforms under the aggregation platform by setting priority dispatching platform rules and dispatching score calculation mechanisms, forming a dispatching queue and executing dispatches sequentially. However, existing methods still have room for optimization in balancing compliance constraints and matching mechanism design. Summary of the Invention

[0003] The purpose of this invention is to propose a two-layer matching-based order dispatching method for ride-hailing aggregation platforms. This method includes the aggregation platform receiving and broadcasting orders, sub-platforms partially filtering candidate vehicles, the aggregation platform making a final global selection from the candidate set, and finally dispatching the order and providing feedback. This invention effectively reduces computational and communication complexity while ensuring information isolation and compliance, thereby improving the overall matching efficiency of the cross-platform ride-hailing market.

[0004] Technical Solution: To solve the above-mentioned technical problems, this invention provides a ride-hailing aggregation platform dispatching method based on two-layer matching, which includes the following steps:

[0005] (1) Passenger order distribution: The aggregation platform integrates passenger order information and broadcasts each order only to the set of sub-platforms preferred by that passenger;

[0006] (2) Sub-platform local matching: Each sub-platform performs local matching based on its own fleet, with the goal of maximizing utility;

[0007] (3) Global assignment by the aggregation platform: The aggregation platform summarizes the candidate results of each sub-platform and determines the final service vehicle for the passenger.

[0008] Furthermore, in step (1), the specific process of passenger order distribution is as follows:

[0009] The aggregation platform has Each sub-platform Managing a dedicated fleet of vehicles, referred to as a vehicle group. , The fleets of all sub-platforms together constitute the total vehicle set of the aggregation platform. ,have The fleets of each sub-platform do not overlap, for any All have This indicates any vehicle It can only belong to one sub-platform;

[0010] In a dispatch cycle Within the platform, multiple passengers submit ride-hailing requests, forming a passenger aggregate. For each passenger Its preferred set of sub-platforms is , indicating passengers Only willing to be assigned to vehicles under these sub-platforms, making symbols , They represent passengers respectively. The origin and destination of the order; Indicates the order origin To the finish line The cost of driving; For passengers The order price is determined by both the order's travel distance and travel time; Indicates passenger The order revenue includes:

[0011] ;

[0012] The aggregation platform will gather passengers within this period. All ride-hailing requests are packaged together; each passenger request Including the origin of the order ,end ,income Collection of Preference Sub-platforms Information, denoted as After integration, the aggregation platform will simultaneously broadcast this order information package to The sub-platforms included in it are used to initiate a distributed local matching process.

[0013] Furthermore, in step (2), the specific process of local matching of the sub-platform is as follows:

[0014] definition Matching drivers and passengers The utility of using a vehicle Serving passengers The resulting utility, under given pricing and cost calculation rules, is determined by the pick-up distance and order price, namely:

[0015]

[0016] In the formula, Indicates vehicle Go to the order origin from your current location. The resulting driving costs, i.e., pick-up and drop-off costs;

[0017] Each sub-platform Only orders from passengers whose preferences include the sub-platform will be received; that is, only orders that meet the criteria will be accepted. Passengers The request to record these passengers as a set Each sub-platform Based on its local vehicle collection Calculate vehicles Serving passengers The utility value is calculated, and local matching is performed. This is a bipartite graph matching problem that aims to maximize the total utility of the system. The mathematical model is as follows:

[0018]

[0019]

[0020]

[0021]

[0022] In the formula, A vector of decision variables; For 0-1 decision variables, when its value is 1, it represents a sub-platform. Vehicle Assigned as passenger The candidate service vehicles, and when its value is 0, it indicates that the vehicle is... Not selected as a passenger After the candidate vehicles are calculated, the sub-platform Determine the candidate vehicle set Return to the aggregation platform.

[0023] Furthermore, in step (3), the specific process of global assignment by the aggregation platform is as follows:

[0024] The aggregation platform gathers the candidate vehicle sets returned by all sub-platforms, for any passenger The aggregation platform only integrates its preferred set of sub-platforms. The returned candidate vehicle results constitute the passenger Global candidate vehicle set:

[0025]

[0026] If a passenger's candidate set If the order is not processed, it means that none of the preferred sub-platforms can provide services for the passenger in the current order dispatch cycle. In this case, the aggregation platform will postpone the passenger's order to the next order dispatch cycle for processing.

[0027] Passengers whose candidate sets are not empty constitute a set. For sets For each passenger in the pool, the aggregation platform performs the final matching, with the goal of selecting from a candidate set for each passenger. The vehicle with the highest utility is selected for assignment, and this process is achieved by solving the following optimization model:

[0028]

[0029]

[0030]

[0031] In the formula, A vector of decision variables; This is a 0-1 decision variable; when its value is 1, it indicates that the aggregation platform ultimately decides to allocate the vehicle... Assigned to passengers When its value is 0, it means that the vehicle has not been selected;

[0032] For all satisfying Driver-passenger matching pair The aggregation platform generates dispatch instructions. And distributed to the vehicles under their management. Each sub-platform, after receiving the instruction, will then push the order dispatch result to the corresponding driver's terminal and wait for the driver's order acceptance confirmation.

[0033] Furthermore, this invention proposes a ride-hailing aggregation platform dispatch system based on two-layer matching, which includes the following modules:

[0034] The passenger order distribution module aggregates passenger order information from the platform and broadcasts each order to a set of sub-platforms preferred by that passenger.

[0035] The sub-platform local matching module allows each sub-platform to perform local matching based on its own fleet, with the goal of maximizing utility.

[0036] The aggregation platform's global assignment module summarizes the candidate results from various sub-platforms and determines the final service vehicle for passengers.

[0037] Furthermore, this invention proposes a non-transitory computer-readable storage medium storing a computer program, which, when executed by a processor, implements any one of the aforementioned two-layer matching-based ride-hailing aggregation platform dispatching methods.

[0038] Furthermore, the present invention proposes an electronic device, including a processor, a storage device, and a computer program stored on the storage device and executable on the processor. When the processor executes the program, the program implements any one of the two-layer matching-based ride-hailing aggregation platform dispatching methods described above.

[0039] Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

[0040] This invention proposes a two-layer matching-based dispatching method for ride-hailing aggregation platforms. While ensuring information isolation and operational compliance, it achieves efficient integration of cross-platform ride-hailing resources by combining local matching on sub-platforms with global coordination on the aggregation platform. This solution not only provides a feasible technical path for multi-platform collaborative scheduling but also lays the foundation for building a fair and orderly urban transportation market. Attached Figure Description

[0041] Figure 1 This is a flowchart of the method of the present invention;

[0042] Figure 2 This is a city traffic network diagram used as an example. Detailed Implementation

[0043] The present invention will be further described below with reference to examples and accompanying drawings.

[0044] This invention provides a method for dispatching ride-hailing services on a two-layer matching platform, such as... Figure 1 As shown, the process includes the following:

[0045] 1) Passenger order distribution: The aggregation platform integrates passenger origin and destination information and broadcasts each order to the set of sub-platforms preferred by that passenger.

[0046] A certain aggregation platform has Each sub-platform. ( Managing a dedicated fleet of vehicles, denoted as a vehicle set. The fleets of all sub-platforms together constitute the total vehicle set of the aggregation platform. ,have The fleets on each sub-platform do not overlap, meaning there are... , , This indicates any vehicle It can only belong to one sub-platform.

[0047] The main business scope of this aggregation platform includes, for example... Figure 2 The urban transportation network shown has 13 nodes and 19 bidirectional road segments. Information on each road segment is shown in the table below, with segment lengths in kilometers.

[0048] Table 1 Road Section Information

[0049]

[0050] In one During the dispatch cycle, there are a total of A number of passengers submitted ride-hailing requests through the aggregation platform. The origin and destination information for each passenger's order is as follows: The passengers are respectively: ( ),passenger ( ),passenger ( The collection of passenger preferences across different sub-platforms ( The following are respectively: , , .

[0051] Assuming the vehicle travels at a constant speed of 30 km / h, the unit driving cost is 0.767 yuan / km. (Order) price The calculation method is as follows: a base price of 12 yuan, a distance-based price of 2.23 yuan / km, and a time-based price of 0.45 yuan / minute. For long-distance orders, an additional 0.75 yuan per kilometer is charged for distances between 20-35 kilometers, and an additional 0.90 yuan per kilometer is charged for distances exceeding 35 kilometers. The order origin is determined based on Dijkstra's algorithm. To the finish line The shortest path between them is used to calculate the travel cost, which is the length of the shortest path multiplied by the unit travel cost. The order price is then calculated by combining the pricing rules mentioned above. With order revenue As shown in the table below:

[0052] Table 2 Passenger Order Information

[0053]

[0054] The aggregation platform will gather passengers within this period. All ride-hailing requests are packaged together; each passenger request Including the origin of the order ,end ,income Collection of Preference Sub-platforms Key information, etc., are recorded as After integration, the aggregation platform will simultaneously broadcast this order information package to [the relevant authorities / platforms]. The sub-platforms included in it are used to initiate a distributed local matching process.

[0055] 2) Sub-platform local matching. Each sub-platform performs local matching based on its own fleet, with the goal of maximizing utility.

[0056] Sub-platforms 1, 2, and 3 each have three vehicles, with the following specific configurations:

[0057] Sub-platform 1: The initial positions are nodes 2, 6, and 10, respectively.

[0058] Sub-platform 2: The initial positions are nodes 5, 9, and 13, respectively.

[0059] Sub-platform 3: The initial positions are nodes 3, 7, and 11 in sequence.

[0060] Define vehicle Serving passengers The effect produced Its value is the revenue from that passenger order minus the vehicle's revenue. Go to the order origin from your current location. Pick-up cost ,Right now:

[0061]

[0062] Among them, the cost of picking up the driver The calculation method is as follows: based on the road segment network, Dijkstra's algorithm is applied to obtain vehicle data. Current location to order origin The shortest path length is multiplied by the unit travel cost.

[0063] Each sub-platform Only orders from passengers whose preferences include the sub-platform will be received; that is, only orders that meet the criteria will be accepted. Passengers The request to record these passengers as a set For the three sub-platforms in this example, there are respectively: , , .

[0064] Sub-platform Based on its local vehicle collection Calculate vehicles Serving passengers The shortest pick-up route and pick-up cost obtained ,utility The results are shown in the table below:

[0065] Table 3 Driver-passenger matching information

[0066]

[0067] Sub-platform Based on local vehicle collection Gathering with passengers The mathematical model for bipartite graph matching is as follows:

[0068]

[0069]

[0070]

[0071]

[0072] After calculation, the candidate vehicle sets for sub-platforms 1, 2, and 3 are as follows:

[0073] Sub-platform 1: Driver-passenger matching results are Candidate vehicle set ;

[0074] Sub-platform 2: Driver-passenger matching results are Candidate vehicle set ;

[0075] Sub-platform 3: Driver-passenger matching results are Candidate vehicle set .

[0076] After the calculation is completed, the sub-platform Determine the candidate vehicle set Return to the aggregation platform.

[0077] 3) Global assignment by the aggregation platform. The aggregation platform summarizes the candidate results from each sub-platform and determines the final service vehicle for the passenger.

[0078] For passengers The aggregation platform summarizes the candidate matching results returned by each sub-platform and merges them to generate a global candidate vehicle set for each passenger. ,as follows:

[0079] The utility of matching pairs are respectively , ;

[0080] The utility of matching pairs are respectively , ;

[0081] The utility of matching pairs are respectively , .

[0082] Passengers whose candidate sets are not empty constitute a set. For sets For each passenger in the pool, the aggregation platform performs the final matching, with the goal of selecting from a candidate set for each passenger. The vehicle with the highest utility is selected for assignment. This process is achieved by solving the following optimization model:

[0083]

[0084]

[0085]

[0086] The solution result is: Passengers Matching vehicles on sub-platform 2 ,passenger Matching vehicles on sub-platform 3 ,passenger Matching vehicles of sub-platform 1 The final set of driver-passenger matching pairs is as follows:

[0087]

[0088] For all satisfying Driver-passenger matching pair The aggregation platform generates dispatch instructions. And distributed to the vehicles under their management. Each sub-platform, after receiving the instruction, will then push the order dispatch result to the corresponding driver's terminal and wait for the driver's order acceptance confirmation.

[0089] The above embodiments are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make several improvements and equivalent substitutions without departing from the principle of the present invention. All such improvements and equivalent substitutions to the claims of the present invention fall within the protection scope of the present invention.

Claims

1. A ride-hailing aggregation platform dispatching method based on two-layer matching, characterized in that, The method includes the following steps: (1) Passenger order distribution: The aggregation platform integrates passenger order information and broadcasts each order only to the set of sub-platforms preferred by that passenger; (2) Sub-platform local matching: Each sub-platform performs local matching based on its own fleet, with the goal of maximizing utility; (3) Global assignment by the aggregation platform: The aggregation platform summarizes the candidate results of each sub-platform and determines the final service vehicle for the passenger.

2. The ride-hailing aggregation platform dispatching method based on two-layer matching according to claim 1, characterized in that, In step (1), the specific process of passenger order distribution is as follows: The aggregation platform has Each sub-platform Managing a dedicated fleet of vehicles, referred to as a vehicle group. , The fleets of all sub-platforms together constitute the total vehicle set of the aggregation platform. ,have The fleets of each sub-platform do not overlap, for any All have This indicates any vehicle It can only belong to one sub-platform; In a dispatch cycle Within the platform, multiple passengers submit ride-hailing requests, forming a passenger aggregate. For each passenger Its preferred set of sub-platforms is , indicating passengers Only willing to be assigned to vehicles under these sub-platforms, making symbols , They represent passengers respectively. The origin and destination of the order; Indicates the order origin To the finish line The cost of driving; For passengers Order price; Indicates passenger Order revenue: ； The aggregation platform will gather passengers within this period. All ride-hailing requests are packaged together; each passenger The request includes the origin of the order. ,end ,income Collection of Preference Sub-platforms Information, denoted as After integration, the aggregation platform will simultaneously broadcast this order information package to The sub-platforms included in it are used to initiate a distributed local matching process.

3. The ride-hailing aggregation platform dispatching method based on two-layer matching according to claim 2, characterized in that, In step (2), the specific process of local matching of the sub-platform is as follows: definition Matching drivers and passengers The utility of using a vehicle Serving passengers The resulting utility, under given pricing and cost calculation rules, is determined by the pick-up distance and order price, namely: ； In the formula, Indicates vehicle Go to the order origin from your current location. The resulting driving costs, i.e., pick-up and drop-off costs; Each sub-platform Only orders from passengers whose preferences include the sub-platform will be received; that is, only orders that meet the criteria will be accepted. Passengers The request to record these passengers as a set Each sub-platform Based on its local vehicle collection Calculate vehicles Serving passengers The utility value is calculated, and local matching is performed. This is a bipartite graph matching problem that aims to maximize the total utility of the system. The mathematical model is as follows: ； ； ； ； In the formula, For decision variable vectors; For 0-1 decision variables, when its value is 1, it represents a sub-platform. Vehicle Assigned as passenger The candidate service vehicles; when its value is 0, it indicates that the vehicle is... Not selected as a passenger After the candidate vehicles are calculated, the sub-platform Determine the candidate vehicle set Return to the aggregation platform.

4. The ride-hailing aggregation platform dispatching method based on two-layer matching according to claim 3, characterized in that, In step (3), the specific process of global assignment by the aggregation platform is as follows: The aggregation platform gathers the candidate vehicle sets returned by all sub-platforms, for any passenger The aggregation platform only integrates its preferred set of sub-platforms. The returned candidate vehicle results constitute the passenger Global candidate vehicle set: ； If a passenger's candidate set If the order is not processed, it means that none of the preferred sub-platforms can provide services for the passenger in the current order dispatch cycle. In this case, the aggregation platform will postpone the passenger's order to the next order dispatch cycle for processing. Passengers whose candidate sets are not empty constitute a set. For sets For each passenger in the pool, the aggregation platform performs the final matching, with the goal of selecting from a candidate set for each passenger. The vehicle with the highest utility is selected for assignment, and this process is achieved by solving the following optimization model: ； ； ； In the formula, For decision variable vectors; This is a 0-1 decision variable; when its value is 1, it indicates that the aggregation platform ultimately decides to allocate the vehicle... Assigned to passengers When its value is 0, it means that the vehicle has not been selected; For all satisfying Driver-passenger matching pair The aggregation platform generates dispatch instructions. And distributed to the vehicles under their management. Each sub-platform, upon receiving an instruction, pushes the order dispatch result to the corresponding driver's terminal and waits for the driver's order acceptance confirmation.

5. A ride-hailing aggregation platform dispatching system based on two-layer matching, characterized in that, The system includes the following modules: The passenger order distribution module aggregates passenger order information from the platform and broadcasts each order to a set of sub-platforms preferred by that passenger. The sub-platform local matching module allows each sub-platform to perform local matching based on its own fleet, with the goal of maximizing utility. The aggregation platform's global assignment module summarizes the candidate results from various sub-platforms and determines the final service vehicle for passengers.

6. A non-transitory computer-readable storage medium, characterized in that, The readable storage medium stores a computer program that, when executed by a processor, implements the ride-hailing aggregation platform dispatching method based on two-layer matching as described in any one of claims 1-4.

7. An electronic device comprising a processor, a storage device, and a computer program stored in the storage device and executable on the processor, characterized in that, When the processor executes the program, the program implements the ride-hailing aggregation platform dispatching method based on two-layer matching as described in any one of claims 1-4.

Images