Order task matching optimization method and system suitable for online car-hailing platform

Abstract

The application relates to the technical field of data processing of a car-hailing platform, in particular to an order task matching optimization method and system suitable for an online car-hailing platform. According to the distribution of the pickup locations of passengers in different historical moments in different car-using areas and the number of passengers, the regional scheduling possibility of each car-using area is obtained; for the latest moment, according to the distribution characteristics of the pickup locations of each passenger in each car-using area and the starting locations of different drivers in adjacent car-using areas and the preset driving time, and the regional scheduling possibility of each car-using area, the driving difficulty of each passenger in each car-using area is obtained; based on the preset historical travel information of passengers and the driving difficulty of each passenger in each car-using area in the latest moment, the car-using demand degree of each car-using area is obtained, and the order matching of passengers and drivers is optimized. The car-using demand degree of each car-using area is accurately obtained, and the order matching of drivers and passengers is optimized.

Description

Technical Field

[0001] This invention relates to the field of data processing technology for ride-hailing platforms, specifically to a method and system for optimizing order task matching on online ride-hailing platforms. Background Technology

[0002] Online ride-hailing platforms collect a large amount of data in real time, including user orders, driver locations, traffic conditions, and historical orders. They use shortest path algorithms to track the locations of drivers and passengers in real time and combine this with real-time traffic flow data to match users with vehicles. Due to the influence of urban planning and layout, traffic demand and vehicle supply vary significantly in different parts of the city at different times.

[0003] In existing technologies, online ride-hailing platforms typically employ regional segmentation and dispatch strategies based on nearest distance or simple weights. However, regional segmentation limits the dispatch range of vehicles, resulting in vehicles being unable to meet the local transportation needs in areas with concentrated demand, while idle vehicle resources in other areas are not fully utilized, leading to poor order matching. Summary of the Invention

[0004] To address the technical problems of traditional regional segmentation limiting vehicle dispatch range and resulting in poor order task matching, the present invention aims to provide an order task matching optimization method and system suitable for online ride-hailing platforms. The specific technical solution adopted is as follows:

[0005] This invention proposes an order task matching optimization method applicable to online ride-hailing platforms, the method comprising:

[0006] Obtain the passenger's ride-hailing location and destination location, as well as the driver's starting location at each moment;

[0007] Based on the distribution of passengers' ride-hailing locations and destination locations at different historical times, multiple ride-hailing areas are obtained; based on the distribution of passengers' ride-hailing locations and the number of passengers in each ride-hailing area at different times of the day, the daily ride-hailing load and ride-hailing density of each ride-hailing area are obtained; based on the ride-hailing load and ride-hailing density of different ride-hailing areas on different historical days, the regional dispatching possibility of each ride-hailing area is obtained.

[0008] For the latest time, based on the location distribution characteristics of each passenger's ride-hailing location in each ride-hailing area and the starting locations of different drivers in adjacent ride-hailing areas, as well as the preset travel time and the regional dispatch possibility of each ride-hailing area, the travel difficulty of each passenger in each ride-hailing area is obtained.

[0009] Based on preset passenger historical travel information, the probability of car demand status for each car-hailing area within a preset future time period is obtained; based on the probability of car demand status for each car-hailing area and the driving difficulty of each passenger in each car-hailing area at the latest moment, the degree of car demand for each car-hailing area is obtained, and the order matching between passengers and drivers is optimized.

[0010] Furthermore, the method for obtaining the vehicle usage area includes:

[0011] Obtain the first relative distance between the two passengers' taxi-hailing locations, obtain the second relative distance between the two passengers' destination locations, and obtain the average of the first and second relative distances as the average distance between the two passengers;

[0012] Based on the average distance between different passengers at different historical moments, DBSCAN clustering is performed on all passengers to obtain multiple passenger clusters. The range formed by the starting and destination positions of all passengers in a cluster is a vehicle usage area.

[0013] Furthermore, the method for obtaining the vehicle load level includes:

[0014] For any given day, the sum of the number of passengers in each usage area at all times within each hour is obtained. If it exceeds the preset passenger threshold, the corresponding hour is designated as the peak period. The sum of the durations of all consecutive adjacent peak periods is compared with the total duration of the previous day to determine the daily usage load level.

[0015] Furthermore, the method for obtaining vehicle usage density includes:

[0016] For any given day, obtain the cumulative number of passengers for each vehicle usage area at all times during all peak hours, and calculate the ratio of the cumulative number of passengers to the area of ​​the vehicle usage area as the number of passengers per unit area.

[0017] The ratio of the number of passengers per unit area to the vehicle load is used as the daily vehicle usage density for each usage area.

[0018] Furthermore, the method for obtaining the regional scheduling possibility includes:

[0019] Based on the vehicle load and density of each vehicle usage area on different historical days, the vehicle demand variation factor for each vehicle usage area on each historical day is obtained.

[0020] Obtain the DTW distance of the sequence of vehicle demand change factors on different historical days between each vehicle use area and adjacent vehicle use areas, as the regional vehicle demand difference; obtain the mean of the vehicle demand change factor of each vehicle use area on all historical days, as the benchmark change factor of vehicle demand; calculate the sum of the ratio of the difference of the benchmark change factor of vehicle demand between each vehicle use area and all adjacent vehicle use areas to the regional vehicle demand difference, as the first sum value.

[0021] Obtain the average vehicle density of each usage area over all historical days as the overall vehicle density; obtain the ratio of the first cumulative value of each usage area to the overall vehicle density, and perform a negative correlation mapping as the area scheduling probability of each usage area.

[0022] Furthermore, the method for obtaining the vehicle demand change factor includes:

[0023] Based on the differences in vehicle load levels between historical days and adjacent days for each vehicle usage area, as well as the daily vehicle usage density, the vehicle demand change factor for each vehicle usage area on historical days is obtained. Both the differences in load levels and the daily vehicle usage density are positively correlated with the vehicle demand change factor.

[0024] Furthermore, the method for obtaining the driving difficulty includes:

[0025] Obtain the preset driving distance between the pick-up location of each passenger in each ride-hailing area and the starting location of each driver in adjacent ride-hailing areas;

[0026] The average ratio of the driving distance to the preset driving time for all drivers in adjacent vehicle use areas is obtained as the driving speed. The driving speed is negatively correlated and the product of the negative correlation mapping result and the regional scheduling probability of each vehicle use area is calculated as the driving difficulty for each passenger in each vehicle use area.

[0027] Furthermore, the method for obtaining the probability of the vehicle usage demand state includes:

[0028] Based on the preset historical travel information of passengers, the probability of the demand for vehicles in each area within a preset time period is obtained by using a Markov chain model.

[0029] Furthermore, the method for obtaining the degree of vehicle demand includes:

[0030] The product of the probability of the demand status in each vehicle area and the driving difficulty of each passenger in each vehicle area at the latest moment is obtained and used as the first demand coefficient of each passenger in each vehicle area.

[0031] The invention obtains the cumulative value of the first ride-hailing demand coefficient of all passengers in each ride-hailing area, and normalizes it to represent the ride-hailing demand level of each ride-hailing area. The invention also proposes an order task matching optimization system suitable for online ride-hailing platforms, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the steps of the order task matching optimization method for online ride-hailing platforms described above.

[0032] The present invention has the following beneficial effects:

[0033] This invention obtains multiple ride-hailing areas based on the distribution of passengers' hail and destination locations at different historical times. Historical hail and destination locations reveal naturally formed demand hotspots, allowing for the division of similar areas. Considering the significant differences in peak traffic demand times and vehicle supply across different areas, the invention obtains the regional dispatch probability for each ride-hailing area based on the distribution of passengers' hail locations and the number of passengers at different historical times, reflecting the possibility of dispatching to the ride-hailing area from adjacent areas. For the latest time, the invention further divides the ride-hailing areas based on the hail locations of each passenger within each ride-hailing area and the starting positions of different drivers in adjacent ride-hailing areas. This invention obtains the travel difficulty for each passenger within each travel area by analyzing the characteristics of the area, preset travel time, and the regional dispatch probability of each travel area, reflecting the degree of difficulty of passenger travel. Based on preset historical passenger travel information, it obtains the probability of travel demand status for each travel area within a preset future time period, considering both the regularity of historical regional data and the influence of regional vehicle data on historical status, reflecting the travel demand situation of each travel area. Based on the probability of travel demand status for each travel area and the travel difficulty for each passenger within each travel area at the latest moment, it obtains the degree of travel demand for each travel area, optimizing passenger and driver order matching. This invention optimizes driver and passenger order matching by accurately obtaining the degree of travel demand for each travel area. Attached Figure Description

[0034] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 A flowchart illustrating an order task matching optimization method applicable to online ride-hailing platforms, provided as an embodiment of the present invention;

[0036] Figure 2This is a flowchart illustrating a method for obtaining the possibility of regional scheduling, as provided in one embodiment of the present invention. Detailed Implementation

[0037] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an order task matching optimization method and system applicable to online ride-hailing platforms proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0039] The following description, in conjunction with the accompanying drawings, details a specific solution for an order task matching optimization method and system applicable to online ride-hailing platforms provided by the present invention.

[0040] Please see Figure 1 The diagram illustrates a flowchart of an order task matching optimization method for online ride-hailing platforms, provided by an embodiment of the present invention, specifically including:

[0041] Step S1: Obtain the passenger's ride-hailing location and destination location, as well as the driver's starting location at each time point.

[0042] In the embodiments of the present invention, considering that the traditional vehicle order matching based on regional segmentation limits the vehicle dispatch range, and that the imbalance between regional traffic demand and vehicle supply during peak hours causes problems such as empty or overloaded vehicles, the analysis considers factors such as passenger location and travel demand. First, real-time data streams are obtained from modules such as the order system, dispatch system, and GPS positioning service through API interfaces to receive high-concurrency real-time data, including: passenger's ride-hailing location and destination location, and driver's starting location.

[0043] It should be noted that, in the embodiments of the present invention, the interval between moments is set by the implementer according to the specific circumstances. In one embodiment of the present invention, it is set to 1 minute. After the missing information is supplemented, data cleaning and quality verification are performed. The specific means are well known to those skilled in the art and will not be described in detail here.

[0044] Step S2: Based on the distribution of passengers' ride-hailing locations and destination locations at different historical times, obtain multiple ride-hailing areas; based on the distribution of passengers' ride-hailing locations and the number of passengers in each ride-hailing area at different times of the day, obtain the daily ride-hailing load and ride-hailing density of each ride-hailing area; based on the ride-hailing load and ride-hailing density of different ride-hailing areas on different historical days, obtain the regional dispatch possibility of each ride-hailing area.

[0045] The location distribution of different orders varies greatly. In the process of matching vehicles and order demand based on the city as a whole, vehicles need to drive long distances empty, resulting in a lot of wasted time and fuel and increased operating costs. Therefore, it is necessary to divide the city based on historical passenger travel data so as to increase the number of drivers while meeting passenger demand and minimizing operating costs. Multiple ride-hailing areas can be obtained based on the distribution of passenger ride-hailing locations and destination locations at different historical times.

[0046] Preferably, in one embodiment of the present invention, the method for obtaining the vehicle usage area includes:

[0047] Obtain the first relative distance between the two passengers' taxi-hailing locations, obtain the second relative distance between the two passengers' destination locations, and obtain the average of the first and second relative distances as the average distance between the two passengers;

[0048] Based on the average distance between different passengers at different historical moments, DBSCAN clustering is performed on all passengers to obtain multiple passenger clusters. The range formed by the starting and destination positions of all passengers in a cluster is a vehicle usage area.

[0049] It should be noted that, in the embodiments of the present invention, the relative distance is calculated by Euclidean distance or Manhattan distance. The closer the average distance between different passengers, the more likely they are to be in the same area. The specific means are well known to those skilled in the art and will not be described in detail here.

[0050] It should be noted that, in the embodiments of the present invention, the specific methods of DBSCAN clustering are well known to those skilled in the art and will not be described in detail here.

[0051] Due to the significant temporal and spatial differences in urban transportation demand and vehicle supply, there are obvious differences in transportation demand and vehicle supply in different regions, resulting in overload or vacancy in vehicle demand in different regions. Therefore, based on the distribution of passenger taxi locations and the number of passengers at different times of the day in each vehicle-using region, we can obtain the daily vehicle load level and vehicle density of each vehicle-using region.

[0052] Preferably, in one embodiment of the present invention, the method for obtaining the vehicle load level includes:

[0053] For any given day, the sum of the number of passengers in each usage area at all times within each hour is obtained. If it exceeds the preset passenger threshold, the corresponding hour is designated as the peak period. The sum of the durations of all consecutive adjacent peak periods is compared with the total duration of the previous day to determine the daily usage load level.

[0054] It should be noted that the more passengers there are, the more likely it is to be a peak time for traffic demand and the greater the demand for vehicles. In the embodiments of the present invention, the method for obtaining the preset passenger threshold based on relevant historical experience is as follows: obtain the sum of the number of passengers at all times in a day and the number of hours of the previous day to obtain the average number of passengers per hour, calculate the average number of passengers per hour multiplied by 1.5, and use it as the preset passenger threshold. If it is greater than the preset passenger threshold, the corresponding hour is a peak period.

[0055] It should be noted that, in the embodiments of the present invention, the time lengths of all consecutive adjacent peak periods are accumulated and the total time length is compared using the same unit, such as using hours, and the total time length in a day is 24 hours.

[0056] Preferably, in one embodiment of the present invention, the method for obtaining vehicle usage density includes:

[0057] For any given day, obtain the cumulative number of passengers for each vehicle usage area at all times during all peak hours, and calculate the ratio of the cumulative number of passengers to the area of ​​the vehicle usage area as the number of passengers per unit area.

[0058] The ratio of the number of passengers per unit area to the vehicle load level is used as the daily vehicle usage density for each usage area.

[0059] Based on this, the higher the cumulative number of passengers at all times during all peak periods, the higher the number of passengers per unit area, the lower the vehicle load, and the shorter the duration of peak periods, the more concentrated the vehicle demand is within the region during the same time period, and the greater the regional load density.

[0060] It should be noted that, in the embodiments of the present invention, the area of ​​the vehicle use area can be calculated according to the map API. The specific means are well known to those skilled in the art and will not be described in detail here.

[0061] Vehicle load reflects the proportion of peak hours each day. The higher the load, the longer the peak hours, and the greater the demand for vehicles. Vehicle density reflects the concentration of passengers. The higher the vehicle density, the greater the demand for vehicles in the usage area, the more vehicles are needed, and the greater the possibility of dispatching. Based on the vehicle load and vehicle density of different usage areas on different historical days, the regional dispatch possibility of each usage area can be obtained.

[0062] Preferably, in one embodiment of the present invention, the method for obtaining the possibility of regional scheduling is described in [reference needed]. Figure 2 It illustrates a flowchart of a method for obtaining the possibility of regional scheduling, including:

[0063] Step S201: Based on the vehicle load and density of each vehicle usage area on different historical days, obtain the vehicle demand variation factor for each vehicle usage area on each historical day.

[0064] Preferably, in one embodiment of the present invention, the method for obtaining the vehicle demand change factor includes:

[0065] Based on the differences in vehicle load levels between historical days and adjacent days for each vehicle usage area, as well as the daily vehicle usage density, the vehicle demand change factor for each vehicle usage area on historical days is obtained. Both the differences in load levels and the daily vehicle usage density are positively correlated with the vehicle demand change factor.

[0066] It should be noted that, in the embodiments of the present invention, the analysis is performed on all historical days based on the current day, excluding the current day.

[0067] It should be noted that the degree difference represents the absolute value of the difference between the calculated vehicle load levels. The greater the degree difference, the more consistent the vehicle load is between adjacent days, resulting in a smaller vehicle demand change factor. The greater the vehicle density, the greater the concentration of vehicle usage, leading to a larger order demand and a larger vehicle demand change factor.

[0068] In one embodiment of the present invention, the product of the degree of difference in vehicle usage intensity between each historical day and the adjacent subsequent day for each vehicle usage area, and the daily vehicle usage density, is obtained as a vehicle usage demand variation factor for each historical day for each vehicle usage area. The formula is expressed as:

[0069] ; Indicates the first The vehicle usage area is in the first Factors affecting daily changes in vehicle usage demand; Indicates the first The vehicle usage area is in the first Daily vehicle usage density; Indicates the first The vehicle usage area is in the first Daily vehicle load level; Indicates the first The vehicle usage area is in the first Daily vehicle load level; This indicates that the absolute value is being calculated.

[0070] Step S202: Obtain the DTW distance of the sequence of vehicle demand change factors for different historical days between each vehicle use area and adjacent vehicle use areas, as the regional vehicle demand difference; obtain the mean of vehicle demand change factors for each vehicle use area over all historical days, as the benchmark change factor of vehicle demand; calculate the sum of the ratio of the difference between the benchmark change factor of vehicle demand and all adjacent vehicle use areas to the regional vehicle demand difference, as the first sum value.

[0071] The greater the DTW distance, the greater the difference in regional vehicle demand, and the more inconsistent the changes in vehicle demand. The larger the first cumulative value, the greater the difference in the baseline change factor of passenger vehicle demand between each vehicle-using area and adjacent areas. A larger baseline change factor of passenger vehicle demand in each area indicates that the change factor of vehicle demand in each area is greater than that in adjacent areas, and that passenger demand in each area is more concentrated and relatively dense. Conversely, the smaller the difference in regional vehicle demand, the more consistent the change factor of vehicle demand, the more concentrated the passenger demand, and the less likely there is to adjust vehicle usage. It should be noted that, considering that the difference in regional vehicle demand may be zero, a manually set non-zero threshold is added when calculating the ratio; its value is specifically set according to the range of values ​​in the denominator.

[0072] Step S203: Obtain the average vehicle density of each vehicle usage area over all historical days as the overall vehicle density; obtain the ratio of the first cumulative value of each vehicle usage area to the overall vehicle density, and perform a negative correlation mapping as the regional scheduling probability of each vehicle usage area.

[0073] It should be noted that, in the embodiments of the present invention, an exponential function with a base of the natural constant is used. A negative correlation mapping is performed. The larger the first cumulative value, the more likely there is a high density of vehicles in the area, and the lower the probability of regional dispatch. The greater the vehicle usage density, the more vehicle usage demand there is in the usage area, the more vehicles are needed, and the greater the probability of dispatch. The specific methods are well known to those skilled in the art and will not be elaborated here.

[0074] In one embodiment of the present invention, the formula for the possibility of regional scheduling is expressed as:

[0075] ;in, Indicates the first The possibility of regional dispatching for each vehicle usage area; Indicates the first The average vehicle density of a vehicle-using area over all historical days, i.e., the overall vehicle density. Indicates the first The average of the factors affecting the change in vehicle demand in each vehicle-using area over all historical days; Indicates the first The adjacent first vehicle use area The average of the factors affecting the change in vehicle demand in each vehicle-using area over all historical days; Indicates the first The number of adjacent vehicle use areas for each vehicle use area; Indicates the first Individual vehicle use areas and adjacent [number] Differences in vehicle usage demand between different vehicle usage areas; This represents an exponential function with the natural constant as its base.

[0076] In the formula for the possibility of regional scheduling, Indicates the calculation of the first Individual vehicle use areas and adjacent [number] The ratio of the difference in the baseline change factor of vehicle demand between different vehicle use areas to the regional differences in vehicle demand is the statistical measure of the difference in vehicle demand. The larger the factor difference, the greater the variation in vehicle demand among different areas. The more concentrated the passenger demand is in a particular vehicle usage area; Indicates obtaining the first The cumulative value of the ratios between a given vehicle usage area and all its adjacent usage areas, i.e., the first cumulative value, indicates that the larger the ratio, the greater the differences and similarities in vehicle usage demand between the areas. The greater the demand for vehicles in a given area, the higher the density of vehicle usage, and the less likely it is that vehicles will be relocated to high-density areas, thus reducing the likelihood of regional dispatching.

[0077] Step S3: For the latest time, based on the location distribution characteristics of each passenger's ride-hailing location in each ride-hailing area and the starting locations of different drivers in adjacent ride-hailing areas, the preset travel time, and the regional dispatch possibility of each ride-hailing area, obtain the travel difficulty of each passenger in each ride-hailing area.

[0078] The path length of each route varies, as does the difficulty of driving. By analyzing the positions of the driver and passenger, we can intuitively understand the distance between the driver and the passenger's location. The greater the distance, the greater the difficulty of driving. For the same distance, the more time it takes, the greater the difficulty of driving. The area dispatch probability reflects the possibility of dispatching a driver from an adjacent area. The lower the probability, the less likely it is that an adjacent area will dispatch a driver to the passenger's area, and the greater the difficulty of driving.

[0079] Preferably, in one embodiment of the present invention, the method for obtaining driving difficulty includes:

[0080] Obtain the preset driving distance between the pick-up location of each passenger in each ride-hailing area and the starting location of each driver in adjacent ride-hailing areas;

[0081] The average ratio of the driving distance to the preset driving time for all drivers in adjacent vehicle use areas is obtained as the driving speed. The driving speed is negatively correlated and the product of the negative correlation mapping result and the regional scheduling probability of each vehicle use area is calculated as the driving difficulty for each passenger in each vehicle use area.

[0082] It should be noted that, in the embodiments of the present invention, the preset driving distance and preset driving time can be obtained in advance by the relevant platform based on historical experience information; through an exponential function with a natural constant as the base. Alternatively, the reciprocal can be used to perform a negative correlation mapping. The specific methods are well-known to those skilled in the art and will not be elaborated here.

[0083] In one embodiment of the present invention, the formula for driving difficulty is expressed as:

[0084] ;in, Indicates the first The first in the vehicle usage area The difficulty of driving for each passenger; Indicates the first [unit] in the adjacent vehicle area The starting position of the driver to the first The first in the vehicle usage area The relative distance between each passenger's taxi pick-up location, i.e., the travel distance; Indicates the first [unit] in the adjacent vehicle area The starting position of the driver to the first The first in the vehicle usage area The preset travel time between each passenger's ride-hailing location; Indicates the first The possibility of regional dispatching for each vehicle usage area; This indicates the number of drivers in adjacent vehicle usage areas; This represents an exponential function with the natural constant as its base.

[0085] In the formula for driving difficulty, This represents the ratio of the driving distance to the preset driving time for drivers within adjacent vehicle usage areas. A larger ratio indicates a greater driving distance, shorter driving time, and higher driving speed. The ratio is calculated using an exponential function with a base of the natural constant to approximate the overall mean. A negative correlation mapping indicates that the lower the driving difficulty, the lower the possibility of regional dispatch, the greater the passenger demand for transportation, and the greater the driving difficulty.

[0086] Step S4: Based on the preset passenger historical travel information, obtain the probability of the demand status of each car-hailing area within the preset time period in the future; based on the probability of the demand status of each car-hailing area and the driving difficulty of each passenger in each car-hailing area at the latest moment, obtain the degree of demand for each car-hailing area and optimize the order matching between passengers and drivers.

[0087] By using a Markov chain model, historical travel patterns are transformed into future demand predictions, enabling data-driven precision operations. In one embodiment of the invention, based on preset passenger historical travel information, a Markov chain model is used to obtain the probability of vehicle demand status for each vehicle-using area within a preset future time period.

[0088] It should be noted that, in one embodiment of the present invention, the travel information of all historical days and the travel information of the latest time are used to analyze the state of the future time period. The preset time period is the range of two hours after the latest time, which is divided into at least the states of needing a car, waiting for a car, and currently taking a car. In other embodiments of the present invention, the preset time period can be set according to specific circumstances, which will not be elaborated here.

[0089] It should be noted that the Markov chain model can predict the probability of the next state by the probability of the previous state. That is, if it is a state of needing a car, the probability of transitioning from the state of needing a car to different states is analyzed, and the sum of the probabilities of needing a car and waiting for a car is calculated to reflect the probability of the state of needing a car. The specific methods are well known to those skilled in the art and will not be elaborated here.

[0090] The higher the probability of a demand for transportation, the greater the demand for transportation in the area. The greater the difficulty of driving, the greater the risk of vehicle transportation, and the greater the demand for transportation. Therefore, based on the probability of demand for transportation in each transportation area and the driving difficulty of each passenger in each transportation area at the latest moment, the degree of demand for transportation in each transportation area can be obtained, and the order matching between passengers and drivers can be optimized.

[0091] Preferably, in one embodiment of the present invention, the method for obtaining the degree of vehicle demand includes:

[0092] The product of the probability of the demand status in each vehicle area and the driving difficulty of each passenger in each vehicle area at the latest moment is obtained and used as the first demand coefficient of each passenger in each vehicle area.

[0093] Obtain the cumulative value of the first car demand coefficient of all passengers in each car use area, and normalize it to represent the car demand level of each car use area.

[0094] It should be noted that, in the embodiments of the present invention, the degree of vehicle demand in each vehicle use area is obtained by using maximum and minimum value normalization, that is, by calculating the cumulative value of the first vehicle demand coefficient of all areas, selecting the maximum and minimum values ​​of the cumulative value, and normalizing the value range to between [0,2]. The specific means are well known to those skilled in the art and will not be described in detail here.

[0095] Based on this, the demand level of all vehicle-using areas is obtained. If the demand level of a vehicle-using area is less than or equal to a preset threshold, the area is designated as a low-density demand area; conversely, it is designated as a high-density demand area. It should be noted that the higher the demand level, the more vehicles are needed for passenger matching, and the lower the demand level, the fewer vehicles are needed for passenger matching. In the embodiments of this invention, the preset threshold is taken as the median value for analysis, i.e., set to 1. A value greater than the preset threshold indicates a higher demand level, and a value less than the preset threshold indicates a lower demand level. To reduce the waiting time for passengers in high-density areas, idle vehicles in adjacent low-density demand areas are dispatched to high-density demand areas to increase traffic demand in high-density demand areas, maximize the satisfaction of urban vehicle demand at different times, balance supply and demand, optimize order matching effects, and significantly improve platform operating efficiency and the experience of both drivers and passengers.

[0096] In summary, this invention obtains multiple ride-hailing areas based on the distribution of passengers' ride-hailing and destination locations at different historical times; it obtains the regional dispatch probability for each ride-hailing area based on the distribution of passengers' ride-hailing locations and the number of passengers in different ride-hailing areas at different historical times; for the latest time, it obtains the driving difficulty for each passenger in each ride-hailing area based on the location distribution characteristics of each passenger's ride-hailing location and the starting locations of different drivers in adjacent ride-hailing areas, the preset travel time, and the regional dispatch probability of each ride-hailing area; based on preset passenger historical travel information, it obtains the probability of ride-hailing demand status for each ride-hailing area in the future preset time period; and combined with the driving difficulty for each passenger in each ride-hailing area at the latest time, it obtains the degree of ride-hailing demand for each ride-hailing area, optimizing passenger and driver order matching. This invention optimizes driver and passenger order matching by accurately obtaining the degree of ride-hailing demand for each ride-hailing area.

[0097] This invention also proposes an order task matching optimization system suitable for online ride-hailing platforms, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any one of the steps of an order task matching optimization method suitable for online ride-hailing platforms.

[0098] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0099] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. An order task matching optimization method suitable for an online car-hailing platform, characterized in that, The method includes: Obtain the passenger's ride-hailing location and destination location, as well as the driver's starting location at each moment; Based on the distribution of passengers' ride-hailing locations and destination locations at different historical times, multiple ride-hailing areas are obtained; based on the distribution of passengers' ride-hailing locations and the number of passengers in each ride-hailing area at different times of the day, the daily ride-hailing load and ride-hailing density of each ride-hailing area are obtained; based on the ride-hailing load and ride-hailing density of different ride-hailing areas on different historical days, the regional dispatching possibility of each ride-hailing area is obtained. For the latest time, based on the location distribution characteristics of each passenger's ride-hailing location in each ride-hailing area and the starting locations of different drivers in adjacent ride-hailing areas, as well as the preset travel time and the regional dispatch possibility of each ride-hailing area, the travel difficulty of each passenger in each ride-hailing area is obtained. Based on the preset passenger historical travel information, the probability of the demand status of each car-hailing area within the preset time period in the future is obtained; based on the probability of the demand status of each car-hailing area and the driving difficulty of each passenger in each car-hailing area at the latest moment, the degree of demand for each car-hailing area is obtained, and the order matching between passengers and drivers is optimized. The method for obtaining the regional scheduling possibility includes: Based on the vehicle load and density of each vehicle usage area on different historical days, the vehicle demand variation factor for each vehicle usage area on each historical day is obtained. Obtain the DTW distance of the sequence of vehicle demand change factors on different historical days between each vehicle use area and adjacent vehicle use areas, as the regional vehicle demand difference; obtain the mean of the vehicle demand change factor of each vehicle use area on all historical days, as the benchmark change factor of vehicle demand; calculate the sum of the ratio of the difference of the benchmark change factor of vehicle demand between each vehicle use area and all adjacent vehicle use areas to the regional vehicle demand difference, as the first sum value. The average vehicle density of each usage area over all historical days is obtained as the overall vehicle density; the ratio of the first cumulative value of each usage area to the overall vehicle density is obtained and negatively correlated to determine the area scheduling probability of each usage area. The method for obtaining the factors of change in vehicle demand includes: Based on the differences in vehicle load levels between historical days and adjacent days for each vehicle usage area, as well as the daily vehicle usage density, the vehicle demand change factor for each vehicle usage area on historical days is obtained. Both the differences in load levels and the daily vehicle usage density are positively correlated with the vehicle demand change factor. 2.The order task matching optimization method suitable for an online car-hailing platform of claim 1, wherein, The method for obtaining the vehicle usage area includes: Obtain the first relative distance between the two passengers' taxi-hailing locations, obtain the second relative distance between the two passengers' destination locations, and obtain the average of the first and second relative distances as the average distance between the two passengers; Based on the average distance between different passengers at different historical moments, DBSCAN clustering is performed on all passengers to obtain multiple passenger clusters. The range formed by the starting and destination positions of all passengers in a cluster is a vehicle usage area. 3.The order task matching optimization method suitable for online car-hailing platforms of claim 1, wherein, The method for obtaining the vehicle load level includes: For any given day, the sum of the number of passengers in each usage area at all times within each hour is obtained. If it exceeds the preset passenger threshold, the corresponding hour is designated as the peak period. The sum of the durations of all consecutive adjacent peak periods is compared with the total duration of the previous day to determine the daily usage load level.

4. The order task matching optimization method applicable to online ride-hailing platforms according to claim 1, characterized in that, The method for obtaining vehicle usage density includes: For any given day, obtain the cumulative number of passengers for each vehicle usage area at all times during all peak hours, and calculate the ratio of the cumulative number of passengers to the area of ​​the vehicle usage area as the number of passengers per unit area. The ratio of the number of passengers per unit area to the vehicle load is used as the daily vehicle usage density for each usage area.

5. The order task matching optimization method suitable for online car-hailing platforms according to claim 1, characterized in that, The methods for obtaining the driving difficulty include: Obtain the preset driving distance between the pick-up location of each passenger in each ride-hailing area and the starting location of each driver in adjacent ride-hailing areas; The average ratio of the driving distance to the preset driving time for all drivers in adjacent vehicle use areas is obtained as the driving speed. The driving speed is negatively correlated and the product of the negative correlation mapping result and the regional scheduling probability of each vehicle use area is calculated as the driving difficulty for each passenger in each vehicle use area.

6. The order task matching optimization method suitable for an online car-hailing platform according to claim 1, characterized in that, The method for obtaining the probability of the vehicle usage demand state includes: Based on the preset historical travel information of passengers, the probability of the demand for vehicles in each area within a preset time period is obtained by using a Markov chain model.

7. The order task matching optimization method suitable for online car-hailing platforms according to claim 1, characterized in that, The methods for obtaining the degree of vehicle demand include: The product of the probability of the demand status in each vehicle area and the driving difficulty of each passenger in each vehicle area at the latest moment is obtained and used as the first demand coefficient of each passenger in each vehicle area. Obtain the cumulative value of the first car demand coefficient of all passengers in each car use area, and normalize it to represent the car demand level of each car use area.

8. An order task matching optimization system suitable for an online ride-hailing platform, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein, When the processor executes the computer program, it implements the steps of the order task matching optimization method applicable to online ride-hailing platforms as described in any one of claims 1 to 7.

Images